Self-improving AI Employees for everyone
Python ai-agentsai-employeeai-employeesassistantautonomous-agentsco-employeescompanygrok-botinstinctloop-agentsmuseown-your-datapersonalself-hosted
feat: Model registry, compaction stats, Telegram integration, UI refinements
trohitgbc537fb20eb3
diff --git a/CLAUDE.md b/CLAUDE.mdindex 583f82ca..6a12e91f 100644--- a/CLAUDE.md+++ b/CLAUDE.md@@ -16,7 +16,7 @@ This is a React Flow-based workflow automation platform implementing n8n-inspire| **[Agent Architecture](./docs-internal/agent_architecture.md)** | How AI Agent and Chat Agent discover skills/tools, inject them into LLM prompts, and execute via LangGraph || **[Agent Delegation](./docs-internal/agent_delegation.md)** | How memory, parameters, and execution context flow when one AI agent delegates work to another agent connected as a tool || **[Agent Teams](./docs-internal/agent_teams.md)** | Claude SDK Agent Teams pattern - AI Employee and Orchestrator nodes with input-teammates handle for multi-agent coordination |-| **[Memory Compaction](./docs-internal/memory_compaction.md)** | Token tracking and memory compaction service using native provider APIs (Anthropic, OpenAI) |+| **[Memory Compaction](./docs-internal/memory_compaction.md)** | Token tracking and model-aware memory compaction using native provider APIs (Anthropic, OpenAI) with threshold = 50% of context window || **[Pricing Service](./docs-internal/pricing_service.md)** | Centralized cost tracking for LLM tokens and API services (Twitter, Google Maps) with HTTPX event hooks || **[CI/CD Pipeline](./docs-internal/ci_cd.md)** | GitHub Actions workflows, predeploy validation, release publishing, and composite setup action || **[Workflow Schema](./docs-internal/workflow-schema.md)** | JSON schema for workflows, edge handle conventions, config node architecture |@@ -63,8 +63,10 @@ server/services/├── node_executor.py # Single node execution with registry pattern├── parameter_resolver.py # Template variable resolution├── agent_team.py # AgentTeamService for multi-agent coordination+├── model_registry.py # ModelRegistryService - model constraints from OpenRouter + llm_defaults├── nodejs_client.py # HTTP client for Node.js code executor├── pricing.py # LLM and API cost calculation (loads config/pricing.json)+├── telegram_service.py # TelegramService with python-telegram-bot long-polling├── tracked_http.py # HTTPX event hooks for automatic API cost tracking├── deployment/ # Event-driven deployment lifecycle│ ├── __init__.py@@ -102,6 +104,7 @@ server/models/server/config/├── llm_defaults.json # Default models per provider (edit to change defaults)+├── model_registry.json # Cached model data from OpenRouter (auto-refreshed)└── pricing.json # LLM and API pricing configserver/nodejs/ # Persistent Node.js server for JS/TS execution@@ -365,15 +368,15 @@ class CacheEntry(SQLModel, table=True):## Codebase Summary- **Hybrid architecture**: Node.js + Python + React TypeScript-- **89 implemented workflow nodes** with clean service separation (6 AI models + 3 AI agents/memory + 13 specialized agents + 11 skills + 3 dedicated tools + 9 dual-purpose tools + 16 Android + 3 WhatsApp + 4 Twitter + 2 Social + 3 Location + 3 Code + 6 Utility + 6 Document + 2 Chat + 2 Scheduler + 1 Workflow + 7 Google Workspace + 1 Apify + 3 Search)-- **WebSocket-First Architecture**: WebSocket as primary frontend-backend communication (87 message handlers)+- **89 implemented workflow nodes** with clean service separation (6 AI models + 3 AI agents/memory + 13 specialized agents + 1 skill + 4 dedicated tools + 3 search + 16 Android + 3 WhatsApp + 4 Twitter + 2 Telegram + 2 Social + 3 Location + 3 Code + 6 Utility + 6 Document + 2 Chat + 2 Scheduler + 2 Workflow + 7 Google Workspace + 1 Apify)+- **WebSocket-First Architecture**: WebSocket as primary frontend-backend communication (119 message handlers)- **Recent optimizations**: REST APIs replaced with WebSocket, AI endpoints migrated to Python, Android automation integrated## Architecture RefactoringThe project was completely refactored from schema-based node definitions to explicit INodeProperties interface system inspired by n8n architecture. Key changes:- **Pure INodeProperties System**: Removed all backward compatibility layers-- **82 Implemented Node Components**: AI models, agents, skills, location services, Android automation, WhatsApp, Twitter/X, social, code execution, HTTP/Webhook utilities, document processing, chat, schedulers, and search-- **WebSocket-First Communication**: 86 WebSocket message handlers replace most REST API calls+- **89 Implemented Node Components**: AI models, agents, skills, location services, Android automation, WhatsApp, Twitter/X, Telegram, social, code execution, HTTP/Webhook utilities, document processing, chat, schedulers, and search+- **WebSocket-First Communication**: 119 WebSocket message handlers replace most REST API calls- **Resource-Operation Pattern**: Organized by functional categories (AI, Location, Android, WhatsApp)- **TypeScript-First**: Full type safety with proper interface alignment- **Code Cleanup**: Removed dead code, unused files, and legacy methods@@ -389,8 +392,8 @@ The project was completely refactored from schema-based node definitions to expl- `src/nodeDefinitions.ts` - Main registry importing and merging all modular node definitions- `src/nodeDefinitions/aiModelNodes.ts` - AI chat model definitions using factory pattern- `src/nodeDefinitions/aiAgentNodes.ts` - AI agent and processing components-- `src/nodeDefinitions/skillNodes.ts` - Skill node definitions (11 nodes including masterSkill and customSkill)-- `src/nodeDefinitions/specializedAgentNodes.ts` - Specialized AI agent definitions (10 nodes) with shared AI_AGENT_PROPERTIES and centralized dracula theming+- `src/nodeDefinitions/skillNodes.ts` - Skill node definition (masterSkill aggregator)+- `src/nodeDefinitions/specializedAgentNodes.ts` - Specialized AI agent definitions (13 nodes) with shared AI_AGENT_PROPERTIES and centralized dracula theming- `src/nodeDefinitions/toolNodes.ts` - AI Agent tool nodes (calculatorTool, currentTimeTool, duckduckgoSearch)- `src/nodeDefinitions/searchNodes.ts` - Search API nodes (braveSearch, serperSearch, perplexitySearch)- `src/nodeDefinitions/androidServiceNodes.ts` - 16 Android service nodes (monitoring, apps, automation, sensors, media)@@ -398,6 +401,7 @@ The project was completely refactored from schema-based node definitions to expl- `src/nodeDefinitions/googleWorkspaceNodes.ts` - All Google Workspace nodes (7 consolidated nodes)- `src/nodeDefinitions/whatsappNodes.ts` - WhatsApp messaging integration (3 nodes)- `src/nodeDefinitions/twitterNodes.ts` - Twitter/X integration (4 nodes: send, search, user, receive)+- `src/nodeDefinitions/telegramNodes.ts` - Telegram bot integration (2 nodes: send, receive)- `src/nodeDefinitions/socialNodes.ts` - Unified social messaging nodes (socialReceive, socialSend)- `src/nodeDefinitions/codeNodes.ts` - Python, JavaScript, and TypeScript code execution nodes- `src/nodeDefinitions/utilityNodes.ts` - HTTP, Webhook, chatTrigger, and console nodes@@ -460,7 +464,7 @@ The theme system uses two complementary color palettes:**Dracula** (vibrant action colors):- `theme.dracula.green` (#50fa7b) - Run/Success actions-- `theme.dracula.purple` (#bd93f9) - Deploy/Save actions+- `theme.dracula.purple` (#bd93f9) - Start/Save actions- `theme.dracula.pink` (#ff79c6) - Cancel/Stop actions- `theme.dracula.cyan` (#8be9fd) - Info/Alternative actions- `theme.dracula.red` (#ff5555) - Error/Danger states@@ -468,7 +472,7 @@ The theme system uses two complementary color palettes:- `theme.dracula.yellow` (#f1fa8c) - Credentials/Highlight#### UI Components Using Dracula Theme-- **TopToolbar Action Buttons**: Run (green), Deploy (purple), Stop (pink), Save (cyan)+- **TopToolbar Action Buttons**: Run (green), Start (purple), Stop (pink), Save (cyan)- **TopToolbar Icon Buttons**: Settings (orange), Credentials (yellow), Theme Toggle (yellow/purple)- **TopToolbar File Button**: Cyan text with medium font size and semibold weight, colorful dropdown menu items- **ParameterPanel Buttons**: Run (green), Save (purple), Cancel (pink)@@ -518,24 +522,24 @@ React Flow edges use dynamic Dracula colors via `getEdgeStyles(theme.dracula)`:### WebSocket-First ArchitectureThe project uses WebSocket as the primary communication method between frontend and backend, replacing most REST API calls:- `src/contexts/WebSocketContext.tsx` - Central WebSocket context with request/response pattern-- `server/routers/websocket.py` - WebSocket endpoint with 87 message handlers+- `server/routers/websocket.py` - WebSocket endpoint with 119 message handlers- `server/services/status_broadcaster.py` - Connection management and broadcasting## Implemented Node Types-The following 108 nodes are currently implemented and functional:+The following 89 nodes are currently implemented and functional:### AI Chat Models (6 nodes)- **openaiChatModel**: OpenAI GPT models with response format options. O-series models (o1, o3, o4) support reasoning effort parameter.- **anthropicChatModel**: Claude models with extended thinking support (budget_tokens for claude-3-5-sonnet, claude-3-opus)- **geminiChatModel**: Google Gemini models with multimodal capabilities, safety settings, and thinking support for 2.5/Flash Thinking models- **openrouterChatModel**: OpenRouter unified API - access 200+ models from OpenAI, Anthropic, Google, Meta, Mistral, and more through a single API. Features free/paid model grouping in dropdown.-- **groqChatModel**: Groq ultra-fast inference with Llama, Mixtral, Gemma, and reasoning-capable Qwen3/QwQ models+- **groqChatModel**: Groq ultra-fast inference with Llama, Qwen3, and GPT-OSS models. Qwen3-32b supports reasoning_format.- **cerebrasChatModel**: Cerebras ultra-fast inference on custom AI hardware with Llama and Qwen models### AI Agents & Memory (3 nodes)- **aiAgent**: Advanced AI agent with tool calling, memory input handle, and iterative reasoning. Uses LangGraph for structured execution. Parameters: Provider, Model, Prompt, System Message, Options.- **chatAgent**: Conversational AI agent with memory and skill support for multi-turn chat interactions. Parameters: Provider, Model, Prompt (supports `{{chatTrigger.message}}` template or auto-fallback from connected input), System Message. Behavior extended by connected skills.-### AI Agent Tool Nodes (5 dedicated + 9 dual-purpose)+### AI Agent Tool Nodes (6 dedicated + 10 dual-purpose)Tool nodes connect to AI Agent's `input-tools` handle to provide capabilities the agent can invoke during reasoning. Both `masterSkill` and `simpleMemory` are in the AI Tools category.#### Dedicated Tool Nodes (passive, tool-only)@@ -723,6 +727,7 @@ if skill_type == 'masterSkill':- **calculatorTool**: Mathematical operations (add, subtract, multiply, divide, power, sqrt, mod, abs)- **currentTimeTool**: Get current date/time with timezone support- **duckduckgoSearch**: DuckDuckGo web search (free, uses `ddgs` library, no API key required)+- **taskManager**: Task management tool for AI agents to create, track, and manage tasks#### Dual-Purpose Search Nodes (workflow node + AI tool)Search API nodes that work BOTH as standalone workflow nodes AND as AI Agent tools. When connected to `input-tools`, the LLM fills the node's parameter schema.@@ -970,6 +975,28 @@ TWITTER_CLIENT_SECRET=your_client_secret# TWITTER_REDIRECT_URI is derived at runtime from request context (no env var needed)```+### Telegram Nodes (2 nodes)+Telegram bot integration using `python-telegram-bot` SDK with long-polling for incoming messages.++- **telegramSend**: **Dual-purpose node** - Send text, photo, document, location, or contact messages via Telegram bot. Works as workflow node OR AI Agent tool. Group: `['social', 'tool']`. Recipient types: Self (bot owner), User/Chat ID, Group. Parameters: recipient_type, chat_id, message_type, text, media_url, caption, parse_mode, silent, reply_to_message_id.+- **telegramReceive**: Event-driven trigger that waits for incoming Telegram messages. Group: `['social', 'trigger']`. Filters: chat type (all/private/group/supergroup/channel), content type (text/photo/video/audio/document/sticker/location/contact/poll), chat ID, user ID, keywords. Option to ignore bot messages.++**Key Files:**+| File | Description |+|------|-------------|+| `server/services/telegram_service.py` | TelegramService with long-polling message handler |+| `server/services/handlers/telegram.py` | Node handlers for send and receive |+| `client/src/nodeDefinitions/telegramNodes.ts` | 2 node definitions |+| `client/src/components/CredentialsModal.tsx` | Telegram bot token panel |++**Authentication:** Bot token from @BotFather on Telegram. Stored via `auth_service.get_api_key('telegram_bot_token')`.++**Environment Variables:**+```bash+# Optional: Set owner chat ID for "Self" recipient type+TELEGRAM_OWNER_CHAT_ID=your_chat_id+```+### Google Workspace Nodes (7 nodes)Consolidated Google Workspace integration with 6 unified operation-based nodes + 1 polling trigger. Each service node uses an `operation` parameter to select the action (e.g., gmail with operation: send/search/read). All services share a single OAuth connection with combined scopes.@@ -1313,8 +1340,8 @@ All scripts in `scripts/` are cross-platform Node.js (Windows, macOS, Linux, WSLSee **[Scripts Reference](./docs-internal/SCRIPTS.md)** for full documentation.## Current Status-✅ **INodeProperties System**: Fully implemented with 75 functional node components-✅ **WebSocket-First Architecture**: 87 message handlers replacing REST APIs+✅ **INodeProperties System**: Fully implemented with 89 functional node components+✅ **WebSocket-First Architecture**: 119 message handlers replacing REST APIs✅ **Code Editor**: Python, JavaScript, and TypeScript executors with syntax-highlighted editor (react-simple-code-editor + prismjs) and console output✅ **Node.js Executor**: Persistent Node.js server (Express + tsx) for fast JS/TS execution, replacing subprocess spawning✅ **Component Palette**: Emoji icons with distinct dracula-themed category colors, localStorage persistence for collapsed sections@@ -1343,7 +1370,7 @@ See **[Scripts Reference](./docs-internal/SCRIPTS.md)** for full documentation.✅ **Production Deployment**: Docker Compose deployment (4 containers: Redis, Backend, Frontend, WhatsApp), nginx reverse proxy, and Let's Encrypt SSL✅ **Authentication System**: n8n-style JWT authentication with HttpOnly cookies, single-owner and multi-user modes✅ **Cache System**: n8n-pattern cache with Redis (production) / SQLite (local dev) / Memory fallback hierarchy-✅ **AI Thinking/Reasoning**: Extended thinking for Claude, Gemini 2.5/Flash Thinking, Groq Qwen3/QwQ with output available in Input Data & Variables for downstream nodes+✅ **AI Thinking/Reasoning**: Extended thinking for Claude, Gemini 2.5/3, OpenAI GPT-5/o-series, Groq Qwen3 with output available in Input Data & Variables for downstream nodes✅ **Onboarding Service**: 5-step welcome wizard with Ant Design UI, database persistence, skip/resume/replay support## Key Features@@ -1624,13 +1651,23 @@ deploy_workflow() -> Sets up triggers, returns immediately- **Execution Engine**: Routes AI nodes to Python Flask backend with auto-injection of API keys#### Supported AI Providers & Models-All 6 providers are available across aiAgent, chatAgent (Zeenie), and all specialized agents:-- **OpenAI**: GPT-5, GPT-4o, o-series (o1, o3, o4) with reasoning effort. Max output: 128K tokens.-- **Anthropic**: Claude Opus 4.6, Sonnet 4.5 with extended thinking. Max output: 128K tokens.-- **Google**: Gemini 3.0, 2.5 Pro/Flash with thinking_budget. Max output: 65K tokens.-- **Groq**: Ultra-fast Llama, Qwen3/QwQ with reasoning_format. Max output: 8K tokens.-- **OpenRouter**: Unified API for 200+ models from multiple providers.-- **Cerebras**: Ultra-fast Llama and Qwen models on custom AI hardware.+All 6 providers are available across aiAgent, chatAgent (Zeenie), and all specialized agents. Model parameters (max output, context length, thinking type, temperature range) are managed by `ModelRegistryService` (`server/services/model_registry.py`) which fetches from OpenRouter and falls back to `server/config/llm_defaults.json`.++| Provider | Key Models | Context | Max Output | Thinking | Temp Range |+|----------|-----------|---------|-----------|----------|------------|+| **OpenAI** | GPT-5.2/5.1/5/5-mini/5-nano | 400K | 128K | effort (hybrid) | 0-2 |+| **OpenAI** | GPT-4.1/4.1-mini/4.1-nano | 1M | 32K | none | 0-2 |+| **OpenAI** | o1, o3, o3-mini, o4-mini | 200K | 100K | effort (reasoning) | fixed 1.0 |+| **OpenAI** | GPT-4o/4o-mini | 128K | 16K | none | 0-2 |+| **Anthropic** | Claude Opus 4.6 | 200K (1M beta) | 128K | budget | 0-1 |+| **Anthropic** | Claude Sonnet 4.6/4.5/4, Opus 4.5, Haiku 4.5 | 200K (1M beta) | 64K | budget | 0-1 |+| **Anthropic** | Claude Opus 4.1/4 | 200K | 32K | budget | 0-1 |+| **Google** | Gemini 3-pro/flash, 2.5-pro/flash/flash-lite | 1M | 65K | budget | 0-2 |+| **Groq** | Llama 4 Scout, Llama 3.x, Qwen3-32b, GPT-OSS | 131K | 8-131K | format (Qwen3) | 0-2 |+| **OpenRouter** | 200+ models from multiple providers | varies | varies | varies | 0-2 |+| **Cerebras** | Llama 3.1-8b, GPT-OSS-120b, Qwen-3-235b | 32-131K | 8K | format (Qwen) | 0-1.5 |++`_resolve_max_tokens()` in `ai.py` clamps user-requested max_tokens to the model's actual limit.#### Key Features- Visual configuration with status indicators and parameter buttons@@ -1645,12 +1682,12 @@ All 6 providers are available across aiAgent, chatAgent (Zeenie), and all speciaUsers can configure default parameter values per LLM provider in the Credentials Modal. These defaults are applied to new AI nodes using that provider.**Configurable Parameters:**-- `temperature` (0-2): Controls randomness in responses-- `max_tokens` (1-128000): Maximum response length+- `temperature`: Controls randomness (range varies: Anthropic 0-1, Cerebras 0-1.5, others 0-2; o-series fixed at 1.0)+- `max_tokens` (1-200000): Maximum response length (clamped to model's actual limit by `_resolve_max_tokens()`)- `thinking_enabled`: Enable extended thinking for supported models-- `thinking_budget` (1024-16000): Token budget for thinking-- `reasoning_effort` (low/medium/high): For OpenAI o-series, Groq-- `reasoning_format` (parsed/hidden): For Groq Qwen models+- `thinking_budget` (1024-16000): Token budget for thinking (Claude, Gemini)+- `reasoning_effort` (low/medium/high): For OpenAI o-series and GPT-5 hybrid reasoning+- `reasoning_format` (parsed/hidden): For Groq Qwen3 models**Database Model** (`server/models/database.py`):```python@@ -1743,12 +1780,14 @@ Extended thinking and reasoning capabilities for supported AI models. When enabl#### Supported Providers & Configuration-| Provider | Models | Parameter | Notes |-|----------|--------|-----------|-------|-| **Claude** | opus-4.6, sonnet-4.5 | `thinkingBudget` (1024-16000 tokens) | Requires `max_tokens > budget_tokens`. Temperature auto-set to 1. |-| **Gemini** | gemini-3.0, gemini-2.5-pro/flash | `thinkingBudget` (token count) | Uses `thinking_budget` API parameter |-| **Groq** | qwen3-32b, qwq-32b | `reasoningFormat` ('parsed' or 'hidden') | 'parsed' returns reasoning, 'hidden' returns only final answer |-| **OpenAI** | o1, o3, o4, GPT-5 series | `reasoningEffort` (minimal/low/medium/high/xhigh) | GPT-5.2 supports xhigh reasoning |+| Provider | Models | Parameter | Thinking Type | Notes |+|----------|--------|-----------|---------------|-------|+| **Claude** | All Claude 4.x/3.5 | `thinkingBudget` (1024-16000 tokens) | budget | Requires `max_tokens > budget_tokens`. Temperature auto-set to 1. |+| **Gemini** | gemini-3.x, gemini-2.5-pro/flash | `thinkingBudget` (token count) | budget | Uses `thinking_budget` API parameter |+| **OpenAI** | o1, o3, o3-mini, o4-mini | `reasoningEffort` (low/medium/high) | effort | Reasoning-only models. Temperature fixed at 1.0. |+| **OpenAI** | GPT-5.2/5.1/5/5-mini/5-nano | `reasoningEffort` (low/medium/high/xhigh) | effort | Hybrid reasoning: can operate with or without thinking. |+| **Groq** | qwen3-32b | `reasoningFormat` ('parsed' or 'hidden') | format | 'parsed' returns reasoning, 'hidden' returns only final answer |+| **Cerebras** | qwen-3-235b | `reasoningFormat` ('parsed' or 'hidden') | format | Same format-based reasoning as Groq Qwen |#### Frontend Parameters (`client/src/factories/baseChatModelFactory.ts`)```typescript@@ -1847,7 +1886,8 @@ const isAI = nodeTypeLower.includes('chatmodel') || aiAgentTypes.includes(nodeTy#### Limitations- **OpenAI o-series**: Reasoning summaries are only available to organizations that have completed verification at platform.openai.com. Without verification, `thinking` will be `null`.- **Claude**: `max_tokens` must be greater than `thinkingBudget`. Temperature is automatically set to 1 when thinking is enabled.-- **Groq**: Only Qwen3 and QwQ models support reasoning. Format 'hidden' suppresses reasoning output.+- **Groq**: Only Qwen3-32b supports reasoning (QwQ removed from Groq). Format 'hidden' suppresses reasoning output.+- **Cerebras**: Qwen-3-235b supports format-based reasoning (same as Groq Qwen).## AI Agent Node Architecture@@ -2809,7 +2849,7 @@ Example workflows use the same format as UI exports:"nodeParameters": {"start_1": { "someParam": "value" }},- "version": "0.0.35"+ "version": "0.0.36"}```@@ -2822,7 +2862,7 @@ Example workflows use the same format as UI exports:| `nodes` | Array of node objects with id, type, position, data || `edges` | Array of edge connections between nodes || `nodeParameters` | Optional map of node_id to parameter objects (saved to DB on import) |-| `version` | App version (e.g., "0.0.35") |+| `version` | App version (e.g., "0.0.36") |### Key Files| File | Description |@@ -2886,7 +2926,7 @@ Multi-step welcome wizard that appears after first launch, guiding users throughSee **[Onboarding Service](./docs-internal/onboarding.md)** for full documentation.### Architecture-- **5-step wizard** using existing `Modal` component + Ant Design `Steps`, `Card`, `Button`, `Typography`, `Tag`+- **5-step wizard** using existing `Modal` component (autoHeight, maxWidth 580px, maxHeight 70vh) + Ant Design `Steps`, `Card`, `Button`, `Typography`, `Tag`- **Database persistence** via `UserSettings.onboarding_completed` + `UserSettings.onboarding_step`- **No new WebSocket handlers** -- reuses `get_user_settings` / `save_user_settings`- **Existing users** auto-skip via migration (`examples_loaded=1` -> `onboarding_completed=1`)@@ -3229,36 +3269,33 @@ CompactionService.record() → Save compaction event to DB### Native Provider APIs+**Threshold strategy:** per-session `custom_threshold` > model-aware threshold (50% of context window) > global default.+**Anthropic SDK (tool_runner):**```python-compaction_control = {- "enabled": True,- "context_token_threshold": 100000-}+# Model-aware: threshold computed from model's context window+compaction_control = svc.anthropic_config(model="claude-opus-4.6", provider="anthropic")+# Returns: {"enabled": True, "context_token_threshold": 500000} (50% of 1M)```**Anthropic Messages API:**```python-{- "betas": ["compact-2026-01-12"],- "context_management": {- "edits": [{- "type": "compact_20260112",- "trigger": {"type": "input_tokens", "value": 100000}- }]- }-}+api_config = svc.anthropic_api_config(model="claude-sonnet-4.5", provider="anthropic")+# Returns: {"betas": ["compact-2026-01-12"], "context_management": {"edits": [...]}}+# Threshold auto-computed from model's context window```**OpenAI:**```python-{"context_management": {"compact_threshold": 100000}}+openai_config = svc.openai_config(model="gpt-5.2", provider="openai")+# Returns: {"context_management": {"compact_threshold": 200000}} (50% of 400K)```### Key Files| File | Description ||------|-------------|-| `server/services/compaction.py` | CompactionService with track(), record(), stats(), configure(), compact_context() methods |+| `server/services/compaction.py` | CompactionService with model-aware thresholds, track(), record(), stats(), configure(), compact_context() methods |+| `server/services/model_registry.py` | ModelRegistryService providing context_length for model-aware threshold computation || `server/services/pricing.py` | PricingService with official pricing registry for all 6 providers || `server/services/ai.py` | `_track_token_usage()` with automatic compaction triggering || `server/models/database.py` | TokenUsageMetric, CompactionEvent, SessionTokenState tables (with cost fields) |@@ -3323,36 +3360,41 @@ from services.compaction import get_compaction_servicesvc = get_compaction_service()-# Get provider-specific config+# Get model-aware provider config (threshold = 50% of context window)+anthropic_cfg = svc.anthropic_config(model="claude-opus-4.6", provider="anthropic")+# threshold: 500000 (50% of 1M context)+openai_cfg = svc.openai_config(model="gpt-5.2", provider="openai")+# threshold: 200000 (50% of 400K context)++# Or override with explicit thresholdanthropic_cfg = svc.anthropic_config(threshold=100000)-openai_cfg = svc.openai_config(threshold=100000)-# Track token usage after AI execution+# Track token usage after AI execution (threshold auto-computed from model)result = await svc.track(session_id="user-123",node_id="agent-1",provider="anthropic",- model="claude-3-5-sonnet",+ model="claude-opus-4.6",usage={"input_tokens": 5000, "output_tokens": 1000, "total_tokens": 6000})-# result: {"total": 6000, "total_cost": 0.021, "threshold": 100000, "needs_compaction": False}+# result: {"total": 6000, "total_cost": 0.021, "threshold": 500000, "needs_compaction": False}# Record compaction event after native API handles itawait svc.record(session_id="user-123",node_id="agent-1",provider="anthropic",- model="claude-3-5-sonnet",- tokens_before=105000,+ model="claude-opus-4.6",+ tokens_before=505000,tokens_after=15000,summary="Compacted conversation summary...")-# Get session statistics-stats = await svc.stats("user-123")-# {"session_id": "user-123", "total": 15000, "threshold": 100000, "count": 1}+# Get session statistics (model-aware threshold when model/provider given)+stats = await svc.stats("user-123", model="claude-opus-4.6", provider="anthropic")+# {"session_id": "user-123", "total": 15000, "threshold": 500000, "count": 1}-# Configure per-session settings+# Configure per-session custom threshold (overrides model-aware threshold)await svc.configure("user-123", threshold=50000, enabled=True)```@@ -3403,9 +3445,11 @@ pricing_info = pricing.get_pricing("anthropic", "claude-3-5-sonnet-20241022")Environment variables in `.env`:```bashCOMPACTION_ENABLED=true # Enable/disable compaction globally-COMPACTION_THRESHOLD=100000 # Default token threshold (min: 10000)+COMPACTION_THRESHOLD=100000 # Global default threshold (min: 10000, used when model context unknown)```+**Threshold priority:** per-session `custom_threshold` > model-aware (50% of context window) > `COMPACTION_THRESHOLD`+### Client-Side CompactionWhen the token threshold is exceeded, the service automatically triggers compaction using the AI service to generate a structured summary:@@ -3488,7 +3532,7 @@ compaction_svc.set_ai_service(container.ai_service())- **5-Section Summary**: Follows Claude Code's structured summary format for continuity- **Automatic Triggering**: Compaction triggered in `_track_token_usage()` when threshold exceeded- **Per-Session State**: Each memory session has independent token tracking and thresholds-- **100K Threshold**: Matches Claude Code's default (configurable per session)+- **Model-Aware Threshold**: 50% of model's context window (e.g., 500K for Claude Opus 4.6 with 1M context, 200K for GPT-5.2 with 400K context). Falls back to global `COMPACTION_THRESHOLD` when model info unavailable. Per-session `custom_threshold` always takes priority.- **Singleton Pattern**: Service accessible via `get_compaction_service()`## API Cost Tracking@@ -3644,6 +3688,7 @@ Tool nodes display execution status via the standard node status system:| calculatorTool | CalculatorSchema | `_execute_calculator()` | Math operations || currentTimeTool | CurrentTimeSchema | `_execute_current_time()` | Date/time with timezone || duckduckgoSearch | DuckDuckGoSearchSchema | `_execute_duckduckgo_search()` | DuckDuckGo web search (free) |+| taskManager | TaskManagerSchema | `_execute_task_manager()` | Task creation and management || braveSearch | BraveSearchSchema | `handle_brave_search()` | Brave Search API web results || serperSearch | SerperSearchSchema | `handle_serper_search()` | Google SERP via Serper API || perplexitySearch | PerplexitySearchSchema | `handle_perplexity_search()` | AI-powered search with citations |@@ -4644,7 +4689,7 @@ def has_real_android_devices(self) -> bool:Updated `client_left` and presence handlers use `has_real_android_devices()` instead of checking total device count.-### WebSocket Message Types (86 Handlers)+### WebSocket Message Types (119 Handlers)#### Request/Response Messages (Client -> Server -> Client)| Category | Message Types |@@ -4659,17 +4704,26 @@ Updated `client_left` and presence handlers use `has_real_android_devices()` ins| **AI Operations** | `execute_ai_node`, `get_ai_models` || **API Keys** | `validate_api_key`, `get_stored_api_key`, `save_api_key`, `delete_api_key` || **Claude OAuth** | `claude_oauth_login`, `claude_oauth_status` |+| **Twitter OAuth** | `twitter_oauth_login`, `twitter_oauth_status`, `twitter_logout` |+| **Google OAuth** | `google_oauth_login`, `google_oauth_status`, `google_logout` |+| **AI Proxy** | `test_ai_proxy` || **Android** | `get_android_devices`, `execute_android_action`, `android_relay_connect`, `android_relay_disconnect`, `android_relay_reconnect` || **Maps** | `validate_maps_key` |-| **WhatsApp** | `whatsapp_status`, `whatsapp_qr`, `whatsapp_send`, `whatsapp_start`, `whatsapp_restart`, `whatsapp_groups`, `whatsapp_group_info`, `whatsapp_chat_history`, `whatsapp_rate_limit_get`, `whatsapp_rate_limit_set`, `whatsapp_rate_limit_stats`, `whatsapp_rate_limit_unpause` |+| **Apify** | `validate_apify_key` |+| **WhatsApp** | `whatsapp_status`, `whatsapp_connected_phone`, `whatsapp_qr`, `whatsapp_send`, `whatsapp_start`, `whatsapp_restart`, `whatsapp_groups`, `whatsapp_group_info`, `whatsapp_chat_history`, `whatsapp_rate_limit_get`, `whatsapp_rate_limit_set`, `whatsapp_rate_limit_stats`, `whatsapp_rate_limit_unpause` |+| **Telegram** | `telegram_connect`, `telegram_disconnect`, `telegram_status`, `telegram_send`, `telegram_reconnect`, `telegram_get_me`, `telegram_get_chat` || **Workflow Storage** | `save_workflow`, `get_workflow`, `get_all_workflows`, `delete_workflow` |-| **Chat Messages** | `send_chat_message`, `get_chat_messages`, `clear_chat_messages`, `save_chat_message` |+| **Chat Messages** | `send_chat_message`, `get_chat_messages`, `clear_chat_messages`, `save_chat_message`, `get_chat_sessions` || **Console/Terminal** | `get_console_logs`, `clear_console_logs`, `get_terminal_logs`, `clear_terminal_logs` || **User Skills** | `get_user_skills`, `get_user_skill`, `create_user_skill`, `update_user_skill`, `delete_user_skill` || **Built-in Skills** | `get_skill_content`, `save_skill_content`, `scan_skill_folder`, `list_skill_folders` || **Memory/Skill Reset** | `clear_memory`, `reset_skill` || **User Settings** | `get_user_settings`, `save_user_settings` || **Provider Defaults** | `get_provider_defaults`, `save_provider_defaults` |+| **Pricing** | `get_pricing_config`, `save_pricing_config` |+| **Usage/Compaction** | `get_api_usage_summary`, `get_compaction_stats`, `configure_compaction`, `get_provider_usage_summary` |+| **Agent Teams** | `create_team`, `get_team`, `get_team_status`, `dissolve_team`, `add_team_task`, `claim_team_task`, `complete_team_task`, `get_team_tasks`, `send_team_message`, `get_team_messages` |+| **Model Registry** | `get_model_constraints`, `refresh_model_registry` |#### Broadcast Messages (Server -> All Clients)| Message Type | Description |diff --git a/client/src/assets/icons/telegram/index.ts b/client/src/assets/icons/telegram/index.tsnew file mode 100644index 00000000..cd19b95d--- /dev/null+++ b/client/src/assets/icons/telegram/index.ts@@ -0,0 +1,19 @@+// Official Telegram icon+// Based on n8n's implementation using SVG files++import telegramSvg from './telegram.svg?raw';++// Convert SVG to data URI for use in node definitions+const svgToDataUri = (svg: string): string => {+ const encoded = encodeURIComponent(svg)+ .replace(/'/g, '%27')+ .replace(/"/g, '%22');+ return `data:image/svg+xml,${encoded}`;+};++export const TELEGRAM_ICON = svgToDataUri(telegramSvg);++// Export raw SVG for React component usage+export const TelegramIcons = {+ telegram: telegramSvg,+};diff --git a/client/src/assets/icons/telegram/telegram.svg b/client/src/assets/icons/telegram/telegram.svgnew file mode 100644index 00000000..e30cd7fe--- /dev/null+++ b/client/src/assets/icons/telegram/telegram.svg@@ -0,0 +1,3 @@+<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24" fill="#0088CC">+ <path d="M11.944 0A12 12 0 0 0 0 12a12 12 0 0 0 12 12 12 12 0 0 0 12-12A12 12 0 0 0 12 0a12 12 0 0 0-.056 0zm4.962 7.224c.1-.002.321.023.465.14a.506.506 0 0 1 .171.325c.016.093.036.306.02.472-.18 1.898-.962 6.502-1.36 8.627-.168.9-.499 1.201-.82 1.23-.696.065-1.225-.46-1.9-.902-1.056-.693-1.653-1.124-2.678-1.8-1.185-.78-.417-1.21.258-1.91.177-.184 3.247-2.977 3.307-3.23.007-.032.014-.15-.056-.212s-.174-.041-.249-.024c-.106.024-1.793 1.14-5.061 3.345-.48.33-.913.49-1.302.48-.428-.008-1.252-.241-1.865-.44-.752-.245-1.349-.374-1.297-.789.027-.216.325-.437.893-.663 3.498-1.524 5.83-2.529 6.998-3.014 3.332-1.386 4.025-1.627 4.476-1.635z"/>+</svg>diff --git a/client/src/components/CredentialsModal.tsx b/client/src/components/CredentialsModal.tsxindex ca6e9eef..3a37383a 100644--- a/client/src/components/CredentialsModal.tsx+++ b/client/src/components/CredentialsModal.tsx@@ -19,13 +19,14 @@ import QRCodeDisplay from './ui/QRCodeDisplay';import ApiKeyInput from './ui/ApiKeyInput';import { useApiKeys, ProviderDefaults, ProviderUsageSummary, APIUsageSummary } from '../hooks/useApiKeys';import { useAppTheme } from '../hooks/useAppTheme';-import { useWhatsAppStatus, useAndroidStatus, useTwitterStatus, useGoogleStatus, useWebSocket, RateLimitConfig, RateLimitStats } from '../contexts/WebSocketContext';+import { useWhatsAppStatus, useAndroidStatus, useTwitterStatus, useGoogleStatus, useTelegramStatus, useWebSocket, RateLimitConfig, RateLimitStats } from '../contexts/WebSocketContext';import { useWhatsApp } from '../hooks/useWhatsApp';import {OpenAIIcon, ClaudeIcon, GeminiIcon, GroqIcon, OpenRouterIcon, CerebrasIcon,} from './icons/AIProviderIcons';import { ApifyIcons } from '../assets/icons/apify';import { SearchIcons } from '../assets/icons/search';+import { TelegramIcons } from '../assets/icons/telegram';// ============================================================================// SERVICE ICONS@@ -54,6 +55,13 @@ const XIcon = () => (<TwitterOutlined style={{ fontSize: 20, color: '#000000' }} />);+const TelegramIcon = () => (+ <div+ dangerouslySetInnerHTML={{ __html: TelegramIcons.telegram }}+ style={{ width: 20, height: 20 }}+ />+);+const GoogleWorkspaceIcon = () => (<GoogleOutlined style={{ fontSize: 20, color: '#4285F4' }} />);@@ -128,6 +136,7 @@ const CATEGORIES: Category[] = [items: [{ id: 'whatsapp_personal', name: 'WhatsApp Personal', placeholder: '', color: '#25D366', desc: 'Connect via QR code pairing', CustomIcon: WhatsAppIcon, isSpecial: true, panelType: 'whatsapp' },{ id: 'twitter', name: 'Twitter/X', placeholder: '', color: '#000000', desc: 'Post tweets, search, user lookup', CustomIcon: XIcon, isSpecial: true, panelType: 'twitter' },+ { id: 'telegram', name: 'Telegram Bot', placeholder: '', color: '#0088CC', desc: 'Send messages via Telegram bot', CustomIcon: TelegramIcon, isSpecial: true, panelType: 'telegram' },],},{@@ -186,6 +195,7 @@ const { validateApiKey, saveApiKey, getStoredApiKey, hasStoredKey, removeApiKey,const androidStatus = useAndroidStatus();const twitterStatus = useTwitterStatus();const googleStatus = useGoogleStatus();+ const telegramStatus = useTelegramStatus();// Tag style helper - consistent theming for status tags@@ -394,6 +404,9 @@ const { validateApiKey, saveApiKey, getStoredApiKey, hasStoredKey, removeApiKey,if (item.panelType === 'gmail') {return { connected: googleStatus.connected, label: googleStatus.connected ? googleStatus.email : 'Not Connected' };}+ if (item.panelType === 'telegram') {+ return { connected: telegramStatus.connected, label: telegramStatus.connected ? `@${telegramStatus.bot_username}` : 'Not Connected' };+ }return null;};@@ -531,6 +544,12 @@ const { validateApiKey, saveApiKey, getStoredApiKey, hasStoredKey, removeApiKey,const [gmailLoading, setGmailLoading] = useState<string | null>(null);const [gmailError, setGmailError] = useState<string | null>(null);+ // Telegram state+ const [telegramToken, setTelegramToken] = useState('');+ const [telegramTokenStored, setTelegramTokenStored] = useState<boolean | null>(null);+ const [telegramLoading, setTelegramLoading] = useState<string | null>(null);+ const [telegramError, setTelegramError] = useState<string | null>(null);+// Rate limit stateconst [rateLimitConfig, setRateLimitConfig] = useState<RateLimitConfig | null>(null);const [rateLimitStats, setRateLimitStats] = useState<RateLimitStats | null>(null);@@ -582,6 +601,15 @@ const { validateApiKey, saveApiKey, getStoredApiKey, hasStoredKey, removeApiKey,}}, [visible, hasStoredKey, getStoredApiKey]);+ // Load Telegram token status on mount+ useEffect(() => {+ if (visible) {+ hasStoredKey('telegram_bot_token').then((has) => {+ setTelegramTokenStored(has);+ });+ }+ }, [visible, hasStoredKey]);+// Twitter handlersconst handleTwitterSaveCredentials = async () => {if (!twitterClientId.trim()) {@@ -711,6 +739,73 @@ const { validateApiKey, saveApiKey, getStoredApiKey, hasStoredKey, removeApiKey,}};+ // Telegram handlers+ const handleTelegramConnect = async () => {+ if (!telegramToken.trim()) {+ setTelegramError('Bot token is required');+ return;+ }+ setTelegramLoading('connect');+ setTelegramError(null);+ try {+ const response = await sendRequest('telegram_connect', { token: telegramToken.trim() });+ if (response.success) {+ setTelegramTokenStored(true);+ } else {+ setTelegramError(response.error || 'Failed to connect');+ }+ } catch (err: any) {+ setTelegramError(err.message || 'Failed to connect');+ } finally {+ setTelegramLoading(null);+ }+ };++ const handleTelegramDisconnect = async () => {+ setTelegramLoading('disconnect');+ setTelegramError(null);+ try {+ const response = await sendRequest('telegram_disconnect', {});+ if (response.success) {+ setTelegramTokenStored(false);+ setTelegramToken('');+ } else {+ setTelegramError(response.error || 'Failed to disconnect');+ }+ } catch (err: any) {+ setTelegramError(err.message || 'Failed to disconnect');+ } finally {+ setTelegramLoading(null);+ }+ };++ const handleTelegramReconnect = async () => {+ setTelegramLoading('reconnect');+ setTelegramError(null);+ try {+ const response = await sendRequest('telegram_reconnect', {});+ if (!response.success) {+ setTelegramError(response.error || 'Failed to reconnect');+ }+ } catch (err: any) {+ setTelegramError(err.message || 'Failed to reconnect');+ } finally {+ setTelegramLoading(null);+ }+ };++ const handleTelegramRefreshStatus = async () => {+ setTelegramLoading('refresh');+ setTelegramError(null);+ try {+ await sendRequest('telegram_status', {});+ } catch (err: any) {+ setTelegramError(err.message || 'Failed to refresh status');+ } finally {+ setTelegramLoading(null);+ }+ };+// Android handlersconst handleAndroidSaveKey = async () => {if (!androidApiKey.trim()) return;@@ -1699,6 +1794,185 @@ const { validateApiKey, saveApiKey, getStoredApiKey, hasStoredKey, removeApiKey,);}+ // Telegram panel+ if (selectedItem.panelType === 'telegram') {+ return (+ <div style={{ padding: theme.spacing.xl, display: 'flex', flexDirection: 'column', flex: 1, minHeight: 0 }}>+ <Descriptions+ title={<Space><TelegramIcon /> Telegram Bot</Space>}+ bordered+ column={1}+ size="small"+ style={{+ marginBottom: theme.spacing.xl,+ background: theme.colors.backgroundAlt,+ borderRadius: theme.borderRadius.md,+ }}+ styles={{+ label: {+ backgroundColor: theme.colors.backgroundPanel,+ color: theme.colors.textSecondary,+ fontWeight: theme.fontWeight.medium,+ },+ content: {+ backgroundColor: theme.colors.background,+ color: theme.colors.text,+ },+ }}+ >+ <Descriptions.Item label="Status">+ <Tag style={getTagStyle(telegramStatus.connected ? 'success' : 'error')}>+ {telegramStatus.connected ? 'Connected' : 'Not Connected'}+ </Tag>+ </Descriptions.Item>+ <Descriptions.Item label="Token">+ <Tag style={getTagStyle(telegramTokenStored ? 'success' : 'error')}>+ {telegramTokenStored === null ? 'Checking...' : telegramTokenStored ? 'Stored' : 'Not configured'}+ </Tag>+ </Descriptions.Item>+ {telegramStatus.connected && telegramStatus.bot_username && (+ <Descriptions.Item label="Bot">+ <span style={{ fontFamily: 'monospace', fontSize: theme.fontSize.sm }}>@{telegramStatus.bot_username}</span>+ {telegramStatus.bot_name && <span style={{ color: theme.colors.textSecondary }}> ({telegramStatus.bot_name})</span>}+ </Descriptions.Item>+ )}+ {telegramStatus.connected && telegramStatus.owner_chat_id && (+ <Descriptions.Item label="Owner">+ <span style={{ fontFamily: 'monospace', fontSize: theme.fontSize.sm }}>{telegramStatus.owner_chat_id}</span>+ </Descriptions.Item>+ )}+ </Descriptions>++ {/* Bot Token Input - Only show if not connected */}+ {!telegramStatus.connected && (+ <div style={{ marginBottom: theme.spacing.xl }}>+ <label style={{+ display: 'block',+ fontSize: theme.fontSize.sm,+ fontWeight: theme.fontWeight.medium,+ color: theme.colors.text,+ marginBottom: theme.spacing.sm,+ }}>+ Bot Token+ </label>+ <Input.Password+ value={telegramToken}+ onChange={(e: React.ChangeEvent<HTMLInputElement>) => setTelegramToken(e.target.value)}+ placeholder="123456789:ABCdefGHIjklMNOpqrSTUvwxYZ"+ style={{+ marginBottom: theme.spacing.md,+ backgroundColor: theme.colors.background,+ borderColor: theme.colors.border,+ color: theme.colors.text,+ fontFamily: 'monospace',+ fontSize: theme.fontSize.sm,+ }}+ />+ <div style={{+ fontSize: theme.fontSize.xs,+ color: theme.colors.textMuted,+ lineHeight: 1.5,+ }}>+ Get a bot token from <span style={{ color: theme.dracula.cyan }}>@BotFather</span> on Telegram.+ After connecting, send any message to your bot to register as the owner.+ </div>+ </div>+ )}++ {telegramError && (+ <Alert type="error" message={telegramError} showIcon style={{ marginBottom: theme.spacing.lg }} />+ )}++ {/* Info box */}+ <div style={{+ padding: theme.spacing.md,+ borderRadius: theme.borderRadius.md,+ backgroundColor: `${theme.dracula.cyan}10`,+ border: `1px solid ${theme.dracula.cyan}30`,+ marginBottom: theme.spacing.xl,+ flex: 1,+ }}>+ <div style={{+ fontSize: theme.fontSize.sm,+ color: theme.colors.textSecondary,+ lineHeight: 1.5,+ }}>+ {telegramStatus.connected ? (+ <>Your Telegram bot is connected and polling for messages. Use Telegram nodes in your workflows.</>+ ) : telegramTokenStored ? (+ <>A bot token is stored. Click Reconnect to connect using the stored token.</>+ ) : (+ <>Enter your bot token from @BotFather to get started.</>+ )}+ </div>+ </div>++ {/* Actions */}+ <div style={{+ display: 'flex',+ gap: theme.spacing.sm,+ justifyContent: 'center',+ paddingTop: theme.spacing.md,+ borderTop: `1px solid ${theme.colors.border}`,+ }}>+ {!telegramStatus.connected ? (+ <>+ <Button+ onClick={handleTelegramConnect}+ loading={telegramLoading === 'connect'}+ disabled={!telegramToken.trim() && !telegramTokenStored}+ style={{+ backgroundColor: `${theme.dracula.green}25`,+ borderColor: `${theme.dracula.green}60`,+ color: theme.dracula.green,+ }}+ >+ Connect+ </Button>+ {telegramTokenStored && (+ <Button+ onClick={handleTelegramReconnect}+ loading={telegramLoading === 'reconnect'}+ style={{+ backgroundColor: `${theme.dracula.purple}25`,+ borderColor: `${theme.dracula.purple}60`,+ color: theme.dracula.purple,+ }}+ >+ Reconnect+ </Button>+ )}+ </>+ ) : (+ <Button+ onClick={handleTelegramDisconnect}+ loading={telegramLoading === 'disconnect'}+ style={{+ backgroundColor: `${theme.dracula.pink}25`,+ borderColor: `${theme.dracula.pink}60`,+ color: theme.dracula.pink,+ }}+ >+ Disconnect+ </Button>+ )}+ <Button+ onClick={handleTelegramRefreshStatus}+ loading={telegramLoading === 'refresh'}+ icon={<ReloadOutlined />}+ style={{+ backgroundColor: `${theme.dracula.cyan}25`,+ borderColor: `${theme.dracula.cyan}60`,+ color: theme.dracula.cyan,+ }}+ >+ Refresh+ </Button>+ </div>+ </div>+ );+ }+// Google Workspace panel (Gmail, Calendar, Drive, Sheets, Tasks, Contacts)if (selectedItem.panelType === 'gmail') {return (diff --git a/client/src/components/onboarding/OnboardingWizard.tsx b/client/src/components/onboarding/OnboardingWizard.tsxindex 1e08cd91..359c6737 100644--- a/client/src/components/onboarding/OnboardingWizard.tsx+++ b/client/src/components/onboarding/OnboardingWizard.tsx@@ -57,28 +57,28 @@ const OnboardingWizard: React.FC<OnboardingWizardProps> = ({ onOpenCredentials,isOpen={isVisible}onClose={skip}title="Welcome Guide"- maxWidth="680px"- maxHeight="85vh"+ maxWidth="580px"+ maxHeight="70vh"+ autoHeight><div style={{display: 'flex',flexDirection: 'column',- height: '100%',- padding: '20px 24px 16px',+ padding: '16px 20px 12px',}}>{/* Progress steps */}<Stepscurrent={currentStep}size="small"items={stepItems}- style={{ marginBottom: 20 }}+ style={{ marginBottom: 16 }}/>{/* Step content */}<div style={{- flex: 1,overflowY: 'auto',- minHeight: 0,+ minHeight: 340,+ maxHeight: 'calc(70vh - 200px)',paddingRight: 4,}}>{renderStep()}@@ -89,8 +89,8 @@ const OnboardingWizard: React.FC<OnboardingWizardProps> = ({ onOpenCredentials,display: 'flex',justifyContent: 'space-between',alignItems: 'center',- paddingTop: 16,- marginTop: 16,+ paddingTop: 12,+ marginTop: 12,borderTop: `1px solid ${theme.colors.border}`,}}>{/* Left: Skip */}diff --git a/client/src/components/onboarding/steps/CanvasStep.tsx b/client/src/components/onboarding/steps/CanvasStep.tsxindex 1d3c38a9..7198fbaa 100644--- a/client/src/components/onboarding/steps/CanvasStep.tsx+++ b/client/src/components/onboarding/steps/CanvasStep.tsx@@ -48,7 +48,7 @@ const CanvasStep: React.FC = () => {<Text style={{ fontSize: 11, color: theme.dracula.orange }} strong>Toolbar</Text><div style={{ flex: 1 }} /><Tag color="orange" style={{ fontSize: 10, margin: 0, lineHeight: '18px' }}>Run</Tag>- <Tag color="purple" style={{ fontSize: 10, margin: 0, lineHeight: '18px' }}>Deploy</Tag>+ <Tag color="purple" style={{ fontSize: 10, margin: 0, lineHeight: '18px' }}>Start</Tag><Tag color="cyan" style={{ fontSize: 10, margin: 0, lineHeight: '18px' }}>Save</Tag></div>diff --git a/client/src/components/parameterPanel/MiddleSection.tsx b/client/src/components/parameterPanel/MiddleSection.tsxindex a1420319..df5a3060 100644--- a/client/src/components/parameterPanel/MiddleSection.tsx+++ b/client/src/components/parameterPanel/MiddleSection.tsx@@ -6,7 +6,7 @@ import ToolSchemaEditor from './ToolSchemaEditor';import MasterSkillEditor from './MasterSkillEditor';import { useAppTheme } from '../../hooks/useAppTheme';import { useAppStore } from '../../store/useAppStore';-import { useWebSocket } from '../../contexts/WebSocketContext';+import { useWebSocket, CompactionStats } from '../../contexts/WebSocketContext';import { nodeDefinitions } from '../../nodeDefinitions';import { INodeTypeDescription, INodeProperties } from '../../types/INodeProperties';import { ExecutionResult } from '../../services/executionService';@@ -86,7 +86,7 @@ const MiddleSection: React.FC<MiddleSectionProps> = ({}) => {const theme = useAppTheme();const { currentWorkflow } = useAppStore();- const { clearMemory, resetSkill, sendRequest } = useWebSocket();+ const { clearMemory, resetSkill, sendRequest, compactionStats: contextCompactionStats, updateCompactionStats } = useWebSocket();const [isConsoleExpanded, setIsConsoleExpanded] = useState(true);const [connectedSkills, setConnectedSkills] = useState<ConnectedSkill[]>([]);const [isSkillsExpanded, setIsSkillsExpanded] = useState(true);@@ -97,19 +97,18 @@ const MiddleSection: React.FC<MiddleSectionProps> = ({const [clearLongTermMemory, setClearLongTermMemory] = useState(false);const [isProcessing, setIsProcessing] = useState(false);- // Compaction state- const [compactionStats, setCompactionStats] = useState<{- session_id: string;- total: number;- threshold: number;- count: number;- } | null>(null);+ // Compaction state - reads from WebSocket context (real-time updates via broadcasts)const [connectedMemorySessionId, setConnectedMemorySessionId] = useState<string | null>(null);const [compactionLoading, setCompactionLoading] = useState(false);const [isEditingThreshold, setIsEditingThreshold] = useState(false);- const [editThresholdValue, setEditThresholdValue] = useState<number>(100000);+ const [editThresholdValue, setEditThresholdValue] = useState<number>(0);const [savingThreshold, setSavingThreshold] = useState(false);+ // Derive compaction stats from WebSocket context for the connected memory session+ const compactionStats: CompactionStats | null = connectedMemorySessionId+ ? contextCompactionStats[connectedMemorySessionId] || null+ : null;+// For Memory nodes: track the connected agent's ID for auto-session displayconst [connectedAgentId, setConnectedAgentId] = useState<string | null>(null);@@ -197,7 +196,6 @@ const MiddleSection: React.FC<MiddleSectionProps> = ({useEffect(() => {if (!isAgentWithSkills || !currentWorkflow) {setConnectedMemorySessionId(null);- setCompactionStats(null);return;}@@ -210,7 +208,6 @@ const MiddleSection: React.FC<MiddleSectionProps> = ({if (!memoryEdge) {setConnectedMemorySessionId(null);- setCompactionStats(null);return;}@@ -228,20 +225,28 @@ const MiddleSection: React.FC<MiddleSectionProps> = ({: nodeId; // nodeId is the agent's IDsetConnectedMemorySessionId(actualSessionId);- // Fetch compaction stats for this session+ // Get agent's model/provider for model-aware threshold computation+ const agentModel = parameters.model || '';+ const agentProvider = parameters.provider || '';++ // Fetch compaction stats for this session (with model-aware threshold)setCompactionLoading(true);const statsResponse = await sendRequest<{session_id: string;total: number;threshold: number;count: number;- }>('get_compaction_stats', { session_id: actualSessionId });+ }>('get_compaction_stats', {+ session_id: actualSessionId,+ model: agentModel,+ provider: agentProvider,+ });- if (statsResponse) {- setCompactionStats({+ if (statsResponse && currentWorkflow?.id) {+ updateCompactionStats(currentWorkflow.id, actualSessionId, {session_id: statsResponse.session_id || actualSessionId,total: statsResponse.total || 0,- threshold: statsResponse.threshold || 100000,+ threshold: statsResponse.threshold,count: statsResponse.count || 0});}@@ -253,7 +258,7 @@ const MiddleSection: React.FC<MiddleSectionProps> = ({};fetchMemorySessionId();- }, [nodeId, isAgentWithSkills, currentWorkflow, sendRequest]);+ }, [nodeId, isAgentWithSkills, currentWorkflow, sendRequest, updateCompactionStats, parameters.model, parameters.provider]);// For Memory nodes: find which AI Agent this memory is connected TO// Used to display the auto-derived session ID@@ -798,9 +803,9 @@ const MiddleSection: React.FC<MiddleSectionProps> = ({<InputNumbersize="small"value={editThresholdValue}- onChange={(v) => setEditThresholdValue(v || 100000)}+ onChange={(v) => setEditThresholdValue(v || compactionStats.threshold)}min={10000}- max={1000000}+ max={2000000}step={10000}style={{ width: 100 }}/>@@ -816,7 +821,9 @@ const MiddleSection: React.FC<MiddleSectionProps> = ({session_id: connectedMemorySessionId,threshold: editThresholdValue});- setCompactionStats(prev => prev ? { ...prev, threshold: editThresholdValue } : null);+ if (compactionStats && connectedMemorySessionId && currentWorkflow?.id) {+ updateCompactionStats(currentWorkflow.id, connectedMemorySessionId, { ...compactionStats, threshold: editThresholdValue });+ }setIsEditingThreshold(false);message.success('Threshold updated');} catch (_err) {diff --git a/client/src/components/ui/Modal.tsx b/client/src/components/ui/Modal.tsxindex fe79735b..05e5660d 100644--- a/client/src/components/ui/Modal.tsx+++ b/client/src/components/ui/Modal.tsx@@ -10,6 +10,8 @@ interface ModalProps {maxWidth?: string;maxHeight?: string;headerActions?: React.ReactNode;+ /** When true, modal height fits content up to maxHeight instead of fixed at maxHeight */+ autoHeight?: boolean;}const Modal: React.FC<ModalProps> = ({@@ -20,6 +22,7 @@ const Modal: React.FC<ModalProps> = ({maxWidth = '500px',maxHeight = '80vh',headerActions,+ autoHeight = false,}) => {const theme = useAppTheme();return (@@ -45,7 +48,7 @@ const Modal: React.FC<ModalProps> = ({boxShadow: '0 25px 50px -12px rgba(0, 0, 0, 0.25)',width: maxWidth,minWidth: maxWidth,- height: maxHeight,+ height: autoHeight ? 'auto' : maxHeight,maxHeight: maxHeight,display: 'flex',flexDirection: 'column',diff --git a/client/src/components/ui/TopToolbar.tsx b/client/src/components/ui/TopToolbar.tsxindex e572b150..ec17a006 100644--- a/client/src/components/ui/TopToolbar.tsx+++ b/client/src/components/ui/TopToolbar.tsx@@ -613,12 +613,12 @@ const TopToolbar: React.FC<TopToolbarProps> = ({e.currentTarget.style.backgroundColor = `${theme.colors.actionDeploy}25`;}}}- title="Deploy workflow continuously"+ title="Start workflow"><svg width="12" height="12" viewBox="0 0 24 24" fill="currentColor"><path d="M12 2C6.48 2 2 6.48 2 12s4.48 10 10 10 10-4.48 10-10S17.52 2 12 2zm-2 15l-5-5 1.41-1.41L10 14.17l7.59-7.59L19 8l-9 9z"/></svg>- Deploy+ Start</button>) : (<button@@ -630,7 +630,7 @@ const TopToolbar: React.FC<TopToolbarProps> = ({onMouseLeave={(e) => {e.currentTarget.style.backgroundColor = `${theme.colors.actionStop}25`;}}- title="Stop deployment"+ title="Stop workflow"><svg width="12" height="12" viewBox="0 0 24 24" fill="currentColor"><rect x="6" y="6" width="12" height="12" rx="1"/>diff --git a/client/src/contexts/WebSocketContext.tsx b/client/src/contexts/WebSocketContext.tsxindex dbb13949..befbd62d 100644--- a/client/src/contexts/WebSocketContext.tsx+++ b/client/src/contexts/WebSocketContext.tsx@@ -112,6 +112,7 @@ export interface TelegramStatus {bot_username: string | null;bot_name: string | null;bot_id: string | null;+ owner_chat_id: number | null;}// WhatsApp Rate Limit types (from Go RPC schema)@@ -208,6 +209,15 @@ export interface NodeParameters {timestamp?: number;}+// Per-session compaction/token usage stats (pushed by backend broadcasts)+export interface CompactionStats {+ session_id: string;+ total: number;+ threshold: number;+ count: number;+ total_cost?: number;+}+export interface FullStatus {android: AndroidStatus;api_keys: Record<string, ApiKeyStatus>;@@ -242,11 +252,15 @@ interface WebSocketContextValue {workflowStatus: WorkflowStatus;deploymentStatus: DeploymentStatus;workflowLock: WorkflowLock;+ compactionStats: Record<string, CompactionStats>; // session_id -> stats (current workflow)// Status gettersgetNodeStatus: (nodeId: string) => NodeStatus | undefined;getApiKeyStatus: (provider: string) => ApiKeyStatus | undefined;getVariable: (name: string) => any;++ // Compaction stats+ updateCompactionStats: (workflowId: string, sessionId: string, stats: CompactionStats) => void;requestStatus: () => void;clearNodeStatus: (nodeId: string) => Promise<void>;clearWhatsAppMessages: () => void;@@ -368,6 +382,7 @@ const defaultTelegramStatus: TelegramStatus = {bot_username: null,bot_name: null,bot_id: null,+ owner_chat_id: null,};const WebSocketContext = createContext<WebSocketContextValue | null>(null);@@ -420,6 +435,8 @@ export const WebSocketProvider: React.FC<{ children: React.ReactNode }> = ({ chiconst [workflowStatus, setWorkflowStatus] = useState<WorkflowStatus>(defaultWorkflowStatus);const [deploymentStatus, setDeploymentStatus] = useState<DeploymentStatus>(defaultDeploymentStatus);const [workflowLock, setWorkflowLock] = useState<WorkflowLock>(defaultWorkflowLock);+ // Per-workflow compaction stats: workflow_id -> session_id -> CompactionStats (n8n pattern)+ const [allCompactionStats, setAllCompactionStats] = useState<Record<string, Record<string, CompactionStats>>>({});const wsRef = useRef<WebSocket | null>(null);const reconnectTimeoutRef = useRef<NodeJS.Timeout | null>(null);@@ -621,6 +638,7 @@ export const WebSocketProvider: React.FC<{ children: React.ReactNode }> = ({ chibot_username: data.bot_username || null,bot_name: data.bot_name || null,bot_id: data.bot_id || null,+ owner_chat_id: data.owner_chat_id ?? null,});}break;@@ -951,6 +969,61 @@ export const WebSocketProvider: React.FC<{ children: React.ReactNode }> = ({ chi}break;+ case 'token_usage_update': {+ // Real-time token usage from backend after each AI execution+ const tokenSessionId = message.session_id;+ const tokenWorkflowId = message.workflow_id || currentWorkflowIdRef.current || '';+ const trackingData = message.data || {};+ if (tokenSessionId && tokenWorkflowId) {+ setAllCompactionStats(prev => {+ const wfStats = prev[tokenWorkflowId] || {};+ const existing = wfStats[tokenSessionId];+ return {+ ...prev,+ [tokenWorkflowId]: {+ ...wfStats,+ [tokenSessionId]: {+ session_id: tokenSessionId,+ total: trackingData.total ?? existing?.total ?? 0,+ threshold: trackingData.threshold ?? existing?.threshold ?? 0,+ count: existing?.count ?? 0,+ total_cost: trackingData.total_cost ?? existing?.total_cost,+ }+ }+ };+ });+ }+ break;+ }++ case 'compaction_completed': {+ // Backend completed compaction - increment count if successful+ const compSessionId = message.session_id;+ if (compSessionId) {+ setAllCompactionStats(prev => {+ const updated = { ...prev };+ for (const wid of Object.keys(updated)) {+ if (updated[wid]?.[compSessionId]) {+ updated[wid] = {+ ...updated[wid],+ [compSessionId]: {+ ...updated[wid][compSessionId],+ count: (updated[wid][compSessionId].count || 0) + (message.success ? 1 : 0),+ total: message.tokens_after ?? updated[wid][compSessionId].total,+ }+ };+ }+ }+ return updated;+ });+ }+ break;+ }++ case 'compaction_starting':+ // Could be used for UI loading indicator in the future+ break;+case 'error':console.error('[WebSocket] Server error:', message.code, message.message);break;@@ -1266,6 +1339,23 @@ export const WebSocketProvider: React.FC<{ children: React.ReactNode }> = ({ chireturn allVariables[currentWorkflowId] || {};}, [allVariables, currentWorkflowId]);+ // Derive current workflow's compaction stats (n8n pattern)+ const compactionStats = useMemo(() => {+ if (!currentWorkflowId) return {};+ return allCompactionStats[currentWorkflowId] || {};+ }, [allCompactionStats, currentWorkflowId]);++ // Update compaction stats for a specific workflow + session (called by MiddleSection on initial fetch)+ const updateCompactionStats = useCallback((workflowId: string, sessionId: string, stats: CompactionStats) => {+ setAllCompactionStats(prev => ({+ ...prev,+ [workflowId]: {+ ...(prev[workflowId] || {}),+ [sessionId]: stats,+ }+ }));+ }, []);+// =========================================================================// Core Request/Response Pattern// =========================================================================@@ -2097,6 +2187,10 @@ export const WebSocketProvider: React.FC<{ children: React.ReactNode }> = ({ chideploymentStatus,workflowLock,+ // Compaction stats (real-time via broadcasts)+ compactionStats,+ updateCompactionStats,+// Status gettersgetNodeStatus,getApiKeyStatus,diff --git a/client/src/factories/baseChatModelFactory.ts b/client/src/factories/baseChatModelFactory.tsindex 225de43f..2b7f9995 100644--- a/client/src/factories/baseChatModelFactory.ts+++ b/client/src/factories/baseChatModelFactory.ts@@ -153,7 +153,7 @@ export const STANDARD_PARAMETERS = {name: 'maxTokens',type: 'number' as const,default: 4096,- typeOptions: { minValue: 1, maxValue: 128000 },+ typeOptions: { minValue: 1, maxValue: 200000 },description: 'Maximum number of tokens to generate'},diff --git a/client/src/hooks/useApiKeys.ts b/client/src/hooks/useApiKeys.tsindex 301c76b6..ae47e970 100644--- a/client/src/hooks/useApiKeys.ts+++ b/client/src/hooks/useApiKeys.ts@@ -49,6 +49,19 @@ export interface ProviderUsageSummary {models: ModelUsageSummary[];}+// Model constraints from model registry+export interface ModelConstraints {+ found: boolean;+ model: string;+ provider: string;+ max_output_tokens: number;+ context_length: number;+ temperature_range: [number, number];+ supports_thinking: boolean;+ thinking_type: 'budget' | 'effort' | 'format' | 'none';+ is_reasoning_model: boolean;+}+// API Usage interfaces (for Twitter, Google Maps, etc.)export interface APIOperationSummary {operation: string;@@ -103,6 +116,9 @@ export interface UseApiKeysResult {// API usage summary (Twitter, etc.)getAPIUsageSummary: (service?: string) => Promise<APIUsageSummary[]>;+ // Model constraints from registry+ getModelConstraints: (model: string, provider: string) => Promise<ModelConstraints>;+// StateisValidating: boolean;validationError: string | null;@@ -401,6 +417,36 @@ export const useApiKeys = (): UseApiKeysResult => {}}, [sendRequest, isConnected]);+ /**+ * Get model constraints from the model registry+ */+ const getModelConstraints = useCallback(async (+ model: string,+ provider: string+ ): Promise<ModelConstraints> => {+ const fallback: ModelConstraints = {+ found: false,+ model,+ provider,+ max_output_tokens: 4096,+ context_length: 128000,+ temperature_range: [0, 2],+ supports_thinking: false,+ thinking_type: 'none',+ is_reasoning_model: false,+ };++ if (!isConnected) return fallback;++ try {+ const response = await sendRequest<ModelConstraints>('get_model_constraints', { model, provider });+ return response || fallback;+ } catch (error) {+ console.warn(`Error fetching model constraints for ${provider}/${model}:`, error);+ return fallback;+ }+ }, [sendRequest, isConnected]);+return {validateApiKey,saveApiKey,@@ -415,6 +461,7 @@ export const useApiKeys = (): UseApiKeysResult => {saveProviderDefaults,getProviderUsageSummary,getAPIUsageSummary,+ getModelConstraints,isValidating,validationError,isConnecteddiff --git a/client/src/nodeDefinitions/aiAgentNodes.ts b/client/src/nodeDefinitions/aiAgentNodes.tsindex 00fa923a..8f823f78 100644--- a/client/src/nodeDefinitions/aiAgentNodes.ts+++ b/client/src/nodeDefinitions/aiAgentNodes.ts@@ -146,7 +146,7 @@ export const aiAgentNodes: Record<string, INodeTypeDescription> = {name: 'maxTokens',type: 'number',default: 4096,- typeOptions: { minValue: 1, maxValue: 128000 },+ typeOptions: { minValue: 1, maxValue: 200000 },description: 'Maximum number of tokens to generate'},{@@ -315,7 +315,7 @@ export const aiAgentNodes: Record<string, INodeTypeDescription> = {name: 'maxTokens',type: 'number',default: 4096,- typeOptions: { minValue: 1, maxValue: 128000 },+ typeOptions: { minValue: 1, maxValue: 200000 },description: 'Maximum number of tokens to generate'},{diff --git a/client/src/nodeDefinitions/aiModelNodes.ts b/client/src/nodeDefinitions/aiModelNodes.tsindex 41822c68..df454037 100644--- a/client/src/nodeDefinitions/aiModelNodes.ts+++ b/client/src/nodeDefinitions/aiModelNodes.ts@@ -85,7 +85,7 @@ const claudeConfig: ChatModelConfig = {{...STANDARD_PARAMETERS.maxTokens,default: 4096, // Higher default for Claude to accommodate thinking mode- typeOptions: { minValue: 1, maxValue: 64000 },+ typeOptions: { minValue: 1, maxValue: 128000 },description: 'Maximum tokens to generate. Must be greater than Thinking Budget when thinking is enabled.'},{diff --git a/client/src/nodeDefinitions/specializedAgentNodes.ts b/client/src/nodeDefinitions/specializedAgentNodes.tsindex 50f63009..363a6e8f 100644--- a/client/src/nodeDefinitions/specializedAgentNodes.ts+++ b/client/src/nodeDefinitions/specializedAgentNodes.ts@@ -126,7 +126,7 @@ export const AI_AGENT_PROPERTIES: INodeProperties[] = [name: 'maxTokens',type: 'number',default: 4096,- typeOptions: { minValue: 1, maxValue: 128000 },+ typeOptions: { minValue: 1, maxValue: 200000 },description: 'Maximum number of tokens to generate'},{diff --git a/client/src/nodeDefinitions/telegramNodes.ts b/client/src/nodeDefinitions/telegramNodes.tsindex d6439175..fedfdf5d 100644--- a/client/src/nodeDefinitions/telegramNodes.ts+++ b/client/src/nodeDefinitions/telegramNodes.ts@@ -54,14 +54,28 @@ export const telegramNodes: Record<string, INodeTypeDescription> = {],properties: [// ===== RECIPIENT =====+ {+ displayName: 'Send To',+ name: 'recipient_type',+ type: 'options',+ options: [+ { name: 'Self (Bot Owner)', value: 'self' },+ { name: 'User / Chat ID', value: 'user' },+ { name: 'Group', value: 'group' }+ ],+ default: 'self',+ description: 'Send to bot owner, specific user, or group'+ },{displayName: 'Chat ID',name: 'chat_id',type: 'string',default: '',- required: true,placeholder: '123456789 or @username',- description: 'Telegram chat ID (numeric) or @username'+ description: 'Telegram chat ID (numeric) or @username',+ displayOptions: {+ show: { recipient_type: ['user', 'group'] }+ }},// ===== MESSAGE TYPE =====diff --git a/client/src/services/executionService.ts b/client/src/services/executionService.tsindex b757ee7a..76161919 100644--- a/client/src/services/executionService.ts+++ b/client/src/services/executionService.ts@@ -9,6 +9,7 @@ import { TWITTER_NODE_TYPES } from '../nodeDefinitions/twitterNodes';import { GOOGLE_WORKSPACE_NODE_TYPES } from '../nodeDefinitions/googleWorkspaceNodes';import { APIFY_NODE_TYPES } from '../nodeDefinitions/apifyNodes';import { SEARCH_NODE_TYPES } from '../nodeDefinitions/searchNodes';+import { TELEGRAM_NODE_TYPES } from '../nodeDefinitions/telegramNodes';import { Node, Edge } from 'reactflow';import { INodeExecutionData } from '../types/INodeProperties';import { API_CONFIG } from '../config/api';@@ -235,7 +236,9 @@ export class ExecutionService {// Apify web scraping...APIFY_NODE_TYPES,// Search API nodes- ...SEARCH_NODE_TYPES+ ...SEARCH_NODE_TYPES,+ // Telegram Nodes+ ...TELEGRAM_NODE_TYPES];return supportedTypes.includes(nodeType);diff --git a/client/src/utils/parameterSanitizer.ts b/client/src/utils/parameterSanitizer.tsindex 817d11ca..6b6b7577 100644--- a/client/src/utils/parameterSanitizer.ts+++ b/client/src/utils/parameterSanitizer.ts@@ -44,7 +44,10 @@ const SENSITIVE_SUBSTRINGS = ['authtoken', 'auth_token',];-// eslint-disable-next-line @typescript-eslint/no-unused-vars+/**+ * Check if a parameter key holds sensitive credential data.+ * Returns true if the key should be stripped from exports.+ */function isSensitiveKey(key: string): boolean {if (SAFE_KEYS.has(key)) return false;if (SENSITIVE_EXACT_KEYS.has(key)) return true;@@ -52,12 +55,41 @@ function isSensitiveKey(key: string): boolean {return SENSITIVE_SUBSTRINGS.some(sub => lower.includes(sub));}+// Runtime/state keys that should not appear in exported workflows.+// These are transient execution state, not user configuration.+const RUNTIME_KEYS = new Set([+ 'memoryContent', // Conversation history (personal data)+ 'token_usage', // Execution token metrics+ 'execution_time', // Runtime timing+ 'last_execution', // Last execution result+ 'last_result', // Cached result+]);+/**- * Deep-strip sensitive keys from a parameter object.+ * Deep-strip sensitive and runtime keys from a parameter object.* Returns a new object with sensitive values removed.* Recurses into nested objects but passes arrays through unchanged.*/export function sanitizeParameters(params: Record<string, any>): Record<string, any> {- // TODO: Re-enable sanitization once safe-key allowlist is fully validated- return { ...params };+ const result: Record<string, any> = {};++ for (const [key, value] of Object.entries(params)) {+ // Skip sensitive credential keys+ if (isSensitiveKey(key)) continue;++ // Skip runtime/state keys+ if (RUNTIME_KEYS.has(key)) continue;++ // Recurse into nested objects (but not arrays or null)+ if (value && typeof value === 'object' && !Array.isArray(value)) {+ const cleaned = sanitizeParameters(value);+ if (Object.keys(cleaned).length > 0) {+ result[key] = cleaned;+ }+ } else {+ result[key] = value;+ }+ }++ return result;}diff --git a/docs-internal/memory_compaction.md b/docs-internal/memory_compaction.mdindex d0c77238..3a01b391 100644--- a/docs-internal/memory_compaction.md+++ b/docs-internal/memory_compaction.md@@ -5,7 +5,7 @@The compaction service enables automatic memory compaction, token tracking, and **cost calculation** for MachinaOs specialized agents. It uses a hybrid approach leveraging native provider APIs (Anthropic, OpenAI) when available, with comprehensive token and cost tracking for all providers.**Inspired by:** Claude Code's compaction pattern from the Anthropic SDK-**Default Threshold:** 100,000 tokens (configurable per session)+**Threshold Strategy:** per-session `custom_threshold` > model-aware (50% of context window) > global default (100K)**Cost Calculation:** Official pricing from each provider (per 1M tokens)## Architecture@@ -57,12 +57,18 @@ When using Anthropic's SDK with `tool_runner`, pass the compaction control confifrom services.compaction import get_compaction_servicesvc = get_compaction_service()++# Model-aware threshold (50% of context window)+compaction_config = svc.anthropic_config(model="claude-opus-4.6", provider="anthropic")+# Returns: {"enabled": True, "context_token_threshold": 500000} (50% of 1M context)++# Or override with explicit thresholdcompaction_config = svc.anthropic_config(threshold=100000)# Returns: {"enabled": True, "context_token_threshold": 100000}# Use with tool_runnerresult = await tool_runner(- model="claude-3-5-sonnet-20241022",+ model="claude-opus-4.6",messages=messages,tools=tools,compaction_control=compaction_config@@ -74,21 +80,16 @@ result = await tool_runner(For direct Messages API usage with the compaction beta:```python+# Model-aware threshold+api_config = svc.anthropic_api_config(model="claude-sonnet-4.5", provider="anthropic")+# Threshold auto-computed from model's context window (50% of 1M = 500000)++# Or override with explicit thresholdapi_config = svc.anthropic_api_config(threshold=100000)-# Returns:-# {-# "betas": ["compact-2026-01-12"],-# "context_management": {-# "edits": [{-# "type": "compact_20260112",-# "trigger": {"type": "input_tokens", "value": 100000}-# }]-# }-# }# Use with Messages APIresponse = await client.messages.create(- model="claude-3-5-sonnet-20241022",+ model="claude-sonnet-4.5",messages=messages,**api_config)@@ -99,6 +100,11 @@ response = await client.messages.create(For OpenAI models (including GPT-4 and o-series):```python+# Model-aware threshold+openai_config = svc.openai_config(model="gpt-5.2", provider="openai")+# Returns: {"context_management": {"compact_threshold": 200000}} (50% of 400K)++# Or override with explicit thresholdopenai_config = svc.openai_config(threshold=100000)# Returns: {"context_management": {"compact_threshold": 100000}}```@@ -218,7 +224,7 @@ result = await svc.track(session_id="user-session-123",node_id="agent-node-1",provider="anthropic",- model="claude-3-5-sonnet-20241022",+ model="claude-opus-4.6",usage={"input_tokens": 5000,"output_tokens": 1000,@@ -232,9 +238,12 @@ result = await svc.track(# result:# {# "total": 6000, # New cumulative total-# "threshold": 100000, # Active threshold+# "threshold": 500000, # Model-aware: 50% of 1M context window+# "total_cost": 0.021, # USD cost# "needs_compaction": False# }+#+# Threshold priority: custom_threshold > model-aware (50% context) > global default```### Record Compaction Event@@ -256,13 +265,18 @@ await svc.record(### Get Session Statistics```python-stats = await svc.stats("user-session-123")+# Model-aware threshold when model/provider given+stats = await svc.stats("user-session-123", model="claude-opus-4.6", provider="anthropic")# {# "session_id": "user-session-123",# "total": 15000,-# "threshold": 100000,+# "threshold": 500000, # 50% of 1M context window# "count": 1 # Number of compactions# }++# Without model/provider, falls back to global default+stats = await svc.stats("user-session-123")+# {"session_id": "user-session-123", "total": 15000, "threshold": 100000, "count": 1}```### Configure Per-Session Settings@@ -329,9 +343,15 @@ ws.send(JSON.stringify({```bash# In server/.envCOMPACTION_ENABLED=true # Enable/disable compaction globally (default: true)-COMPACTION_THRESHOLD=100000 # Default token threshold (default: 100000, min: 10000)+COMPACTION_THRESHOLD=100000 # Global fallback threshold (default: 100000, min: 10000)+ # Used when model context window is unknown```+**Threshold priority chain:**+1. Per-session `custom_threshold` (set via `configure()` or WebSocket)+2. Model-aware threshold: 50% of model's context window (e.g., 500K for 1M model)+3. Global `COMPACTION_THRESHOLD` from `.env` (fallback when model info unavailable)+### Per-Session OverrideSessions can override the global threshold:@@ -413,7 +433,8 @@ if memory_data and memory_data.get('session_id'):| File | Description ||------|-------------|-| `server/services/compaction.py` | CompactionService class with provider configs |+| `server/services/compaction.py` | CompactionService class with model-aware thresholds and provider configs |+| `server/services/model_registry.py` | ModelRegistryService providing context_length for threshold computation || `server/models/database.py` | SQLModel tables for token tracking || `server/core/database.py` | CRUD methods for metrics and events || `server/core/config.py` | Environment variable configuration |@@ -429,7 +450,7 @@ if memory_data and memory_data.get('session_id'):3. **Per-Session State**: Each memory session has independent token tracking and thresholds. This allows different agents to have different compaction settings.-4. **100K Token Threshold**: Matches Claude Code's default. This provides a good balance between context retention and performance.+4. **Model-Aware Threshold**: Threshold is 50% of the model's context window (via `get_model_threshold()`). For example, Claude Opus 4.6 (1M context) gets a 500K threshold, GPT-5.2 (400K) gets 200K, Groq models (131K) get ~65K. Falls back to global `COMPACTION_THRESHOLD` when model info is unavailable. Per-session `custom_threshold` always takes priority.5. **Singleton Pattern**: Service accessible via `get_compaction_service()` for easy integration anywhere in the codebase.@@ -437,7 +458,7 @@ if memory_data and memory_data.get('session_id'):## Client-Side Compaction-For all providers (not just Anthropic/OpenAI), the service performs automatic client-side compaction when the threshold is exceeded. This uses the AI service to generate a structured summary following Claude Code's 5-section pattern.+For all providers (not just Anthropic/OpenAI), the service performs automatic client-side compaction when the model-aware threshold is exceeded. This uses the AI service to generate a structured summary following Claude Code's 5-section pattern. The summary max_tokens is capped at `min(4096, model's max output tokens)`.### compact_context() Method@@ -448,7 +469,7 @@ result = await svc.compact_context(memory_content="# Conversation History\n...", # Current memory markdownprovider="anthropic",api_key="sk-...",- model="claude-3-5-sonnet-20241022"+ model="claude-opus-4.6")# result:@@ -604,6 +625,5 @@ The Token Usage panel is displayed in the MiddleSection of the parameter panel f## Future Enhancements-1. **Cost Tracking**: Extend token metrics to include cost calculations per provider/model-2. **Compaction History UI**: View past compaction events and summaries in the frontend-3. **Multiple Summary Strategies**: Allow users to choose different summarization approaches+1. **Compaction History UI**: View past compaction events and summaries in the frontend+2. **Multiple Summary Strategies**: Allow users to choose different summarization approachesdiff --git a/docs-internal/onboarding.md b/docs-internal/onboarding.mdindex ad6e5f37..de9d46ac 100644--- a/docs-internal/onboarding.md+++ b/docs-internal/onboarding.md@@ -135,7 +135,7 @@ export const useOnboarding = (reopenTrigger?: number) => {| `reopenTrigger` | `number?` | Incrementing counter triggers wizard reopen |**UI Structure**:-- Uses existing `Modal` component (`maxWidth="680px"`, `maxHeight="85vh"`)+- Uses existing `Modal` component (`maxWidth="580px"`, `maxHeight="70vh"`, `autoHeight`)- Ant Design `Steps` component for progress indicator- Step content rendered via switch on `currentStep`- Footer: "Skip for now" (left) | "Back" + "Next"/"Start Building" (right)diff --git a/package.json b/package.jsonindex 98450c56..e77ed39e 100644--- a/package.json+++ b/package.json@@ -1,6 +1,6 @@{"name": "machinaos",- "version": "0.0.35",+ "version": "0.0.36","description": "Open source workflow automation platform with AI agents, React Flow, and n8n-inspired architecture","type": "module","keywords": [diff --git a/server/config/llm_defaults.json b/server/config/llm_defaults.jsonindex 99d755b4..4c71716f 100644--- a/server/config/llm_defaults.json+++ b/server/config/llm_defaults.json@@ -1,32 +1,146 @@{"version": "2026.02",- "last_updated": "2026-02-23",- "_description": "Default models and settings for each LLM provider. Edit this file to change defaults.",+ "last_updated": "2026-02-26",+ "_description": "Default models, settings, and offline fallback constraints for each LLM provider.","providers": {"openai": {"default_model": "gpt-5.2",- "detection_patterns": ["gpt", "openai", "o1", "o3", "o4"]+ "detection_patterns": ["gpt", "openai", "o1", "o3", "o4"],+ "max_output_tokens": {+ "gpt-5.2": 128000,+ "gpt-5.1": 128000,+ "gpt-5": 128000,+ "gpt-5-mini": 128000,+ "gpt-5-nano": 128000,+ "gpt-4o": 16384,+ "gpt-4o-mini": 16384,+ "gpt-4.1": 32768,+ "gpt-4.1-mini": 32768,+ "gpt-4.1-nano": 32768,+ "o3": 100000,+ "o3-mini": 100000,+ "o4-mini": 100000,+ "o1": 100000,+ "_default": 16384+ },+ "context_length": {+ "gpt-5.2": 400000,+ "gpt-5.1": 400000,+ "gpt-5": 400000,+ "gpt-5-mini": 400000,+ "gpt-5-nano": 400000,+ "gpt-4.1": 1047576,+ "gpt-4.1-mini": 1047576,+ "gpt-4.1-nano": 1047576,+ "gpt-4o": 128000,+ "gpt-4o-mini": 128000,+ "o3": 200000,+ "o3-mini": 200000,+ "o4-mini": 200000,+ "o1": 200000,+ "_default": 128000+ },+ "reasoning_models": ["o1", "o3", "o3-mini", "o4-mini"],+ "thinking_models": ["gpt-5"],+ "thinking_type": "effort",+ "temperature_range": [0.0, 2.0]},"anthropic": {"default_model": "claude-opus-4.6",- "detection_patterns": ["claude", "anthropic"]+ "detection_patterns": ["claude", "anthropic"],+ "max_output_tokens": {+ "claude-opus-4.6": 128000,+ "claude-sonnet-4.6": 64000,+ "claude-opus-4.5": 64000,+ "claude-sonnet-4.5": 64000,+ "claude-sonnet-4": 64000,+ "claude-opus-4.1": 32000,+ "claude-opus-4": 32000,+ "claude-haiku-4.5": 64000,+ "claude-3.5-haiku": 8192,+ "claude-3-opus": 4096,+ "claude-3-haiku": 4096,+ "_default": 8192+ },+ "context_length": {+ "claude-opus-4.6": 1000000,+ "claude-sonnet-4.6": 1000000,+ "claude-sonnet-4.5": 1000000,+ "claude-sonnet-4": 1000000,+ "_default": 200000+ },+ "thinking_type": "budget",+ "temperature_range": [0.0, 1.0]},"gemini": {"default_model": "gemini-2.5-flash",- "detection_patterns": ["gemini", "google"]+ "detection_patterns": ["gemini", "google"],+ "max_output_tokens": {+ "gemini-3-pro": 65536,+ "gemini-3-flash": 65536,+ "gemini-2.5-pro": 65536,+ "gemini-2.5-flash": 65536,+ "gemini-2.5-flash-lite": 65536,+ "gemini-2.0-flash": 8192,+ "gemini-2.0-flash-lite": 8192,+ "_default": 8192+ },+ "context_length": {+ "_default": 1048576+ },+ "thinking_models": ["gemini-2.5", "gemini-3"],+ "thinking_type": "budget",+ "temperature_range": [0.0, 2.0]},"openrouter": {"default_model": "anthropic/claude-opus-4.6",- "detection_patterns": ["openrouter"]+ "detection_patterns": ["openrouter"],+ "max_output_tokens": {+ "_default": 4096+ },+ "context_length": {+ "_default": 128000+ },+ "temperature_range": [0.0, 2.0]},"groq": {- "default_model": "llama-4-maverick",- "detection_patterns": ["groq"]+ "default_model": "llama-4-scout",+ "detection_patterns": ["groq"],+ "max_output_tokens": {+ "llama-3.1-8b": 131072,+ "llama-3.3-70b": 32768,+ "llama-4-scout": 8192,+ "qwen3-32b": 40960,+ "gpt-oss-120b": 65536,+ "gpt-oss-20b": 65536,+ "_default": 8192+ },+ "context_length": {+ "_default": 131072+ },+ "thinking_models": ["qwen3"],+ "thinking_type": "format",+ "temperature_range": [0.0, 2.0]},"cerebras": {"default_model": "llama3.1-8b",- "detection_patterns": ["cerebras", "llama3.1", "llama-3.3", "qwen-3", "qwen3", "gpt-oss"]+ "detection_patterns": ["cerebras", "llama3.1", "gpt-oss", "qwen-3-235b"],+ "max_output_tokens": {+ "llama3.1-8b": 8192,+ "gpt-oss-120b": 8192,+ "qwen-3-235b": 8192,+ "_default": 8192+ },+ "context_length": {+ "llama3.1-8b": 128000,+ "gpt-oss-120b": 131072,+ "qwen-3-235b": 32768,+ "_default": 128000+ },+ "thinking_models": ["qwen-3-235b"],+ "thinking_type": "format",+ "temperature_range": [0.0, 1.5]}}}diff --git a/server/config/model_registry.json b/server/config/model_registry.jsonnew file mode 100644index 00000000..f676ffdb--- /dev/null+++ b/server/config/model_registry.json@@ -0,0 +1,10337 @@+{+ "_generated": "2026-02-26T17:40:58.078412+00:00",+ "_source": "openrouter",+ "_model_count": 343,+ "models": {+ "gemini/gemini-3.1-flash-image-preview": {+ "id": "google/gemini-3.1-flash-image-preview",+ "name": "Google: Nano Banana 2 (Gemini 3.1 Flash Image Preview)",+ "provider": "gemini",+ "local_id": "gemini-3.1-flash-image-preview",+ "context_length": 65536,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 0.25,+ "output_price_per_mtok": 1.5,+ "supports_thinking": true,+ "thinking_type": "budget",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_p"+ ]+ },+ "qwen/qwen3.5-35b-a3b": {+ "id": "qwen/qwen3.5-35b-a3b",+ "name": "Qwen: Qwen3.5-35B-A3B",+ "provider": "qwen",+ "local_id": "qwen3.5-35b-a3b",+ "context_length": 262144,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 0.25,+ "output_price_per_mtok": 2.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p"+ ]+ },+ "qwen/qwen3.5-27b": {+ "id": "qwen/qwen3.5-27b",+ "name": "Qwen: Qwen3.5-27B",+ "provider": "qwen",+ "local_id": "qwen3.5-27b",+ "context_length": 262144,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 0.3,+ "output_price_per_mtok": 2.4,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p"+ ]+ },+ "qwen/qwen3.5-122b-a10b": {+ "id": "qwen/qwen3.5-122b-a10b",+ "name": "Qwen: Qwen3.5-122B-A10B",+ "provider": "qwen",+ "local_id": "qwen3.5-122b-a10b",+ "context_length": 262144,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 0.4,+ "output_price_per_mtok": 3.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p"+ ]+ },+ "qwen/qwen3.5-flash-02-23": {+ "id": "qwen/qwen3.5-flash-02-23",+ "name": "Qwen: Qwen3.5-Flash",+ "provider": "qwen",+ "local_id": "qwen3.5-flash-02-23",+ "context_length": 1000000,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 0.1,+ "output_price_per_mtok": 0.4,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "liquid/lfm-2-24b-a2b": {+ "id": "liquid/lfm-2-24b-a2b",+ "name": "LiquidAI: LFM2-24B-A2B",+ "provider": "liquid",+ "local_id": "lfm-2-24b-a2b",+ "context_length": 32768,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.03,+ "output_price_per_mtok": 0.12,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "gemini/gemini-3.1-pro-preview-customtools": {+ "id": "google/gemini-3.1-pro-preview-customtools",+ "name": "Google: Gemini 3.1 Pro Preview Custom Tools",+ "provider": "gemini",+ "local_id": "gemini-3.1-pro-preview-customtools",+ "context_length": 1048576,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 2.0,+ "output_price_per_mtok": 12.0,+ "supports_thinking": true,+ "thinking_type": "budget",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "openai/gpt-5.3-codex": {+ "id": "openai/gpt-5.3-codex",+ "name": "OpenAI: GPT-5.3-Codex",+ "provider": "openai",+ "local_id": "gpt-5.3-codex",+ "context_length": 400000,+ "max_output_tokens": 128000,+ "input_price_per_mtok": 1.75,+ "output_price_per_mtok": 14.0,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "tool_choice",+ "tools"+ ]+ },+ "aion-labs/aion-2.0": {+ "id": "aion-labs/aion-2.0",+ "name": "AionLabs: Aion-2.0",+ "provider": "aion-labs",+ "local_id": "aion-2.0",+ "context_length": 131072,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.8,+ "output_price_per_mtok": 1.6,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "temperature",+ "top_p"+ ]+ },+ "gemini/gemini-3.1-pro-preview": {+ "id": "google/gemini-3.1-pro-preview",+ "name": "Google: Gemini 3.1 Pro Preview",+ "provider": "gemini",+ "local_id": "gemini-3.1-pro-preview",+ "context_length": 1048576,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 2.0,+ "output_price_per_mtok": 12.0,+ "supports_thinking": true,+ "thinking_type": "budget",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "anthropic/claude-sonnet-4.6": {+ "id": "anthropic/claude-sonnet-4.6",+ "name": "Anthropic: Claude Sonnet 4.6",+ "provider": "anthropic",+ "local_id": "claude-sonnet-4.6",+ "context_length": 1000000,+ "max_output_tokens": 128000,+ "input_price_per_mtok": 3.0,+ "output_price_per_mtok": 15.0,+ "supports_thinking": true,+ "thinking_type": "budget",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p",+ "verbosity"+ ]+ },+ "qwen/qwen3.5-plus-02-15": {+ "id": "qwen/qwen3.5-plus-02-15",+ "name": "Qwen: Qwen3.5 Plus 2026-02-15",+ "provider": "qwen",+ "local_id": "qwen3.5-plus-02-15",+ "context_length": 1000000,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 0.4,+ "output_price_per_mtok": 2.4,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "qwen/qwen3.5-397b-a17b": {+ "id": "qwen/qwen3.5-397b-a17b",+ "name": "Qwen: Qwen3.5 397B A17B",+ "provider": "qwen",+ "local_id": "qwen3.5-397b-a17b",+ "context_length": 262144,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 0.55,+ "output_price_per_mtok": 3.5,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "minimax/minimax-m2.5": {+ "id": "minimax/minimax-m2.5",+ "name": "MiniMax: MiniMax M2.5",+ "provider": "minimax",+ "local_id": "minimax-m2.5",+ "context_length": 196608,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 0.3,+ "output_price_per_mtok": 1.1,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "min_p",+ "parallel_tool_calls",+ "presence_penalty",+ "reasoning",+ "reasoning_effort",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_logprobs",+ "top_p"+ ]+ },+ "z-ai/glm-5": {+ "id": "z-ai/glm-5",+ "name": "Z.ai: GLM 5",+ "provider": "z-ai",+ "local_id": "glm-5",+ "context_length": 204800,+ "max_output_tokens": 131072,+ "input_price_per_mtok": 0.95,+ "output_price_per_mtok": 2.55,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_logprobs",+ "top_p"+ ]+ },+ "qwen/qwen3-max-thinking": {+ "id": "qwen/qwen3-max-thinking",+ "name": "Qwen: Qwen3 Max Thinking",+ "provider": "qwen",+ "local_id": "qwen3-max-thinking",+ "context_length": 262144,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 1.2,+ "output_price_per_mtok": 6.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "anthropic/claude-opus-4.6": {+ "id": "anthropic/claude-opus-4.6",+ "name": "Anthropic: Claude Opus 4.6",+ "provider": "anthropic",+ "local_id": "claude-opus-4.6",+ "context_length": 1000000,+ "max_output_tokens": 128000,+ "input_price_per_mtok": 5.0,+ "output_price_per_mtok": 25.0,+ "supports_thinking": true,+ "thinking_type": "budget",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p",+ "verbosity"+ ]+ },+ "qwen/qwen3-coder-next": {+ "id": "qwen/qwen3-coder-next",+ "name": "Qwen: Qwen3 Coder Next",+ "provider": "qwen",+ "local_id": "qwen3-coder-next",+ "context_length": 262144,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 0.12,+ "output_price_per_mtok": 0.75,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "openrouter/free": {+ "id": "openrouter/free",+ "name": "Free Models Router",+ "provider": "openrouter",+ "local_id": "free",+ "context_length": 200000,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "stepfun/step-3.5-flash:free": {+ "id": "stepfun/step-3.5-flash:free",+ "name": "StepFun: Step 3.5 Flash (free)",+ "provider": "stepfun",+ "local_id": "step-3.5-flash:free",+ "context_length": 256000,+ "max_output_tokens": 256000,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "stop",+ "temperature",+ "tools",+ "top_p"+ ]+ },+ "stepfun/step-3.5-flash": {+ "id": "stepfun/step-3.5-flash",+ "name": "StepFun: Step 3.5 Flash",+ "provider": "stepfun",+ "local_id": "step-3.5-flash",+ "context_length": 256000,+ "max_output_tokens": 256000,+ "input_price_per_mtok": 0.1,+ "output_price_per_mtok": 0.3,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "stop",+ "temperature",+ "tools",+ "top_p"+ ]+ },+ "arcee-ai/trinity-large-preview:free": {+ "id": "arcee-ai/trinity-large-preview:free",+ "name": "Arcee AI: Trinity Large Preview (free)",+ "provider": "arcee-ai",+ "local_id": "trinity-large-preview:free",+ "context_length": 131000,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "response_format",+ "structured_outputs",+ "temperature",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "moonshotai/kimi-k2.5": {+ "id": "moonshotai/kimi-k2.5",+ "name": "MoonshotAI: Kimi K2.5",+ "provider": "moonshotai",+ "local_id": "kimi-k2.5",+ "context_length": 262144,+ "max_output_tokens": 65535,+ "input_price_per_mtok": 0.45,+ "output_price_per_mtok": 2.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "min_p",+ "parallel_tool_calls",+ "presence_penalty",+ "reasoning",+ "reasoning_effort",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_logprobs",+ "top_p"+ ]+ },+ "upstage/solar-pro-3:free": {+ "id": "upstage/solar-pro-3:free",+ "name": "Upstage: Solar Pro 3 (free)",+ "provider": "upstage",+ "local_id": "solar-pro-3:free",+ "context_length": 128000,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools"+ ]+ },+ "minimax/minimax-m2-her": {+ "id": "minimax/minimax-m2-her",+ "name": "MiniMax: MiniMax M2-her",+ "provider": "minimax",+ "local_id": "minimax-m2-her",+ "context_length": 65536,+ "max_output_tokens": 2048,+ "input_price_per_mtok": 0.3,+ "output_price_per_mtok": 1.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "temperature",+ "top_p"+ ]+ },+ "writer/palmyra-x5": {+ "id": "writer/palmyra-x5",+ "name": "Writer: Palmyra X5",+ "provider": "writer",+ "local_id": "palmyra-x5",+ "context_length": 1040000,+ "max_output_tokens": 8192,+ "input_price_per_mtok": 0.6,+ "output_price_per_mtok": 6.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "liquid/lfm-2.5-1.2b-thinking:free": {+ "id": "liquid/lfm-2.5-1.2b-thinking:free",+ "name": "LiquidAI: LFM2.5-1.2B-Thinking (free)",+ "provider": "liquid",+ "local_id": "lfm-2.5-1.2b-thinking:free",+ "context_length": 32768,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "seed",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "liquid/lfm-2.5-1.2b-instruct:free": {+ "id": "liquid/lfm-2.5-1.2b-instruct:free",+ "name": "LiquidAI: LFM2.5-1.2B-Instruct (free)",+ "provider": "liquid",+ "local_id": "lfm-2.5-1.2b-instruct:free",+ "context_length": 32768,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "seed",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "openai/gpt-audio": {+ "id": "openai/gpt-audio",+ "name": "OpenAI: GPT Audio",+ "provider": "openai",+ "local_id": "gpt-audio",+ "context_length": 128000,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 2.5,+ "output_price_per_mtok": 10.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_logprobs",+ "top_p"+ ]+ },+ "openai/gpt-audio-mini": {+ "id": "openai/gpt-audio-mini",+ "name": "OpenAI: GPT Audio Mini",+ "provider": "openai",+ "local_id": "gpt-audio-mini",+ "context_length": 128000,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 0.6,+ "output_price_per_mtok": 2.4,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_logprobs",+ "top_p"+ ]+ },+ "z-ai/glm-4.7-flash": {+ "id": "z-ai/glm-4.7-flash",+ "name": "Z.ai: GLM 4.7 Flash",+ "provider": "z-ai",+ "local_id": "glm-4.7-flash",+ "context_length": 202752,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.06,+ "output_price_per_mtok": 0.4,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "openai/gpt-5.2-codex": {+ "id": "openai/gpt-5.2-codex",+ "name": "OpenAI: GPT-5.2-Codex",+ "provider": "openai",+ "local_id": "gpt-5.2-codex",+ "context_length": 400000,+ "max_output_tokens": 128000,+ "input_price_per_mtok": 1.75,+ "output_price_per_mtok": 14.0,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "tool_choice",+ "tools"+ ]+ },+ "allenai/molmo-2-8b": {+ "id": "allenai/molmo-2-8b",+ "name": "AllenAI: Molmo2 8B",+ "provider": "allenai",+ "local_id": "molmo-2-8b",+ "context_length": 36864,+ "max_output_tokens": 36864,+ "input_price_per_mtok": 0.2,+ "output_price_per_mtok": 0.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "presence_penalty",+ "repetition_penalty",+ "seed",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "allenai/olmo-3.1-32b-instruct": {+ "id": "allenai/olmo-3.1-32b-instruct",+ "name": "AllenAI: Olmo 3.1 32B Instruct",+ "provider": "allenai",+ "local_id": "olmo-3.1-32b-instruct",+ "context_length": 65536,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.2,+ "output_price_per_mtok": 0.6,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "bytedance-seed/seed-1.6-flash": {+ "id": "bytedance-seed/seed-1.6-flash",+ "name": "ByteDance Seed: Seed 1.6 Flash",+ "provider": "bytedance-seed",+ "local_id": "seed-1.6-flash",+ "context_length": 262144,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.075,+ "output_price_per_mtok": 0.3,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "bytedance-seed/seed-1.6": {+ "id": "bytedance-seed/seed-1.6",+ "name": "ByteDance Seed: Seed 1.6",+ "provider": "bytedance-seed",+ "local_id": "seed-1.6",+ "context_length": 262144,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.25,+ "output_price_per_mtok": 2.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "minimax/minimax-m2.1": {+ "id": "minimax/minimax-m2.1",+ "name": "MiniMax: MiniMax M2.1",+ "provider": "minimax",+ "local_id": "minimax-m2.1",+ "context_length": 196608,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.27,+ "output_price_per_mtok": 0.95,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_logprobs",+ "top_p"+ ]+ },+ "z-ai/glm-4.7": {+ "id": "z-ai/glm-4.7",+ "name": "Z.ai: GLM 4.7",+ "provider": "z-ai",+ "local_id": "glm-4.7",+ "context_length": 202752,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.3,+ "output_price_per_mtok": 1.4,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "gemini/gemini-3-flash-preview": {+ "id": "google/gemini-3-flash-preview",+ "name": "Google: Gemini 3 Flash Preview",+ "provider": "gemini",+ "local_id": "gemini-3-flash-preview",+ "context_length": 1048576,+ "max_output_tokens": 65535,+ "input_price_per_mtok": 0.5,+ "output_price_per_mtok": 3.0,+ "supports_thinking": true,+ "thinking_type": "budget",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "mistralai/mistral-small-creative": {+ "id": "mistralai/mistral-small-creative",+ "name": "Mistral: Mistral Small Creative",+ "provider": "mistralai",+ "local_id": "mistral-small-creative",+ "context_length": 32768,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.1,+ "output_price_per_mtok": 0.3,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "tool_choice",+ "tools"+ ]+ },+ "allenai/olmo-3.1-32b-think": {+ "id": "allenai/olmo-3.1-32b-think",+ "name": "AllenAI: Olmo 3.1 32B Think",+ "provider": "allenai",+ "local_id": "olmo-3.1-32b-think",+ "context_length": 65536,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 0.15,+ "output_price_per_mtok": 0.5,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "xiaomi/mimo-v2-flash": {+ "id": "xiaomi/mimo-v2-flash",+ "name": "Xiaomi: MiMo-V2-Flash",+ "provider": "xiaomi",+ "local_id": "mimo-v2-flash",+ "context_length": 262144,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 0.09,+ "output_price_per_mtok": 0.29,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "nvidia/nemotron-3-nano-30b-a3b:free": {+ "id": "nvidia/nemotron-3-nano-30b-a3b:free",+ "name": "NVIDIA: Nemotron 3 Nano 30B A3B (free)",+ "provider": "nvidia",+ "local_id": "nemotron-3-nano-30b-a3b:free",+ "context_length": 256000,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "seed",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "nvidia/nemotron-3-nano-30b-a3b": {+ "id": "nvidia/nemotron-3-nano-30b-a3b",+ "name": "NVIDIA: Nemotron 3 Nano 30B A3B",+ "provider": "nvidia",+ "local_id": "nemotron-3-nano-30b-a3b",+ "context_length": 262144,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.05,+ "output_price_per_mtok": 0.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "openai/gpt-5.2-chat": {+ "id": "openai/gpt-5.2-chat",+ "name": "OpenAI: GPT-5.2 Chat",+ "provider": "openai",+ "local_id": "gpt-5.2-chat",+ "context_length": 128000,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 1.75,+ "output_price_per_mtok": 14.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "response_format",+ "seed",+ "structured_outputs",+ "tool_choice",+ "tools"+ ]+ },+ "openai/gpt-5.2-pro": {+ "id": "openai/gpt-5.2-pro",+ "name": "OpenAI: GPT-5.2 Pro",+ "provider": "openai",+ "local_id": "gpt-5.2-pro",+ "context_length": 400000,+ "max_output_tokens": 128000,+ "input_price_per_mtok": 21.0,+ "output_price_per_mtok": 168.0,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "tool_choice",+ "tools"+ ]+ },+ "openai/gpt-5.2": {+ "id": "openai/gpt-5.2",+ "name": "OpenAI: GPT-5.2",+ "provider": "openai",+ "local_id": "gpt-5.2",+ "context_length": 400000,+ "max_output_tokens": 128000,+ "input_price_per_mtok": 1.75,+ "output_price_per_mtok": 14.0,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "tool_choice",+ "tools"+ ]+ },+ "mistralai/devstral-2512": {+ "id": "mistralai/devstral-2512",+ "name": "Mistral: Devstral 2 2512",+ "provider": "mistralai",+ "local_id": "devstral-2512",+ "context_length": 262144,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.4,+ "output_price_per_mtok": 2.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "relace/relace-search": {+ "id": "relace/relace-search",+ "name": "Relace: Relace Search",+ "provider": "relace",+ "local_id": "relace-search",+ "context_length": 256000,+ "max_output_tokens": 128000,+ "input_price_per_mtok": 1.0,+ "output_price_per_mtok": 3.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "seed",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "z-ai/glm-4.6v": {+ "id": "z-ai/glm-4.6v",+ "name": "Z.ai: GLM 4.6V",+ "provider": "z-ai",+ "local_id": "glm-4.6v",+ "context_length": 131072,+ "max_output_tokens": 131072,+ "input_price_per_mtok": 0.3,+ "output_price_per_mtok": 0.9,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "nex-agi/deepseek-v3.1-nex-n1": {+ "id": "nex-agi/deepseek-v3.1-nex-n1",+ "name": "Nex AGI: DeepSeek V3.1 Nex N1",+ "provider": "nex-agi",+ "local_id": "deepseek-v3.1-nex-n1",+ "context_length": 131072,+ "max_output_tokens": 163840,+ "input_price_per_mtok": 0.27,+ "output_price_per_mtok": 1.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "response_format",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "essentialai/rnj-1-instruct": {+ "id": "essentialai/rnj-1-instruct",+ "name": "EssentialAI: Rnj 1 Instruct",+ "provider": "essentialai",+ "local_id": "rnj-1-instruct",+ "context_length": 32768,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.15,+ "output_price_per_mtok": 0.15,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "stop",+ "structured_outputs",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "openrouter/bodybuilder": {+ "id": "openrouter/bodybuilder",+ "name": "Body Builder (beta)",+ "provider": "openrouter",+ "local_id": "bodybuilder",+ "context_length": 128000,+ "max_output_tokens": 0,+ "input_price_per_mtok": -1000000.0,+ "output_price_per_mtok": -1000000.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": []+ },+ "openai/gpt-5.1-codex-max": {+ "id": "openai/gpt-5.1-codex-max",+ "name": "OpenAI: GPT-5.1-Codex-Max",+ "provider": "openai",+ "local_id": "gpt-5.1-codex-max",+ "context_length": 400000,+ "max_output_tokens": 128000,+ "input_price_per_mtok": 1.25,+ "output_price_per_mtok": 10.0,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "tool_choice",+ "tools"+ ]+ },+ "amazon/nova-2-lite-v1": {+ "id": "amazon/nova-2-lite-v1",+ "name": "Amazon: Nova 2 Lite",+ "provider": "amazon",+ "local_id": "nova-2-lite-v1",+ "context_length": 1000000,+ "max_output_tokens": 65535,+ "input_price_per_mtok": 0.3,+ "output_price_per_mtok": 2.5,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "mistralai/ministral-14b-2512": {+ "id": "mistralai/ministral-14b-2512",+ "name": "Mistral: Ministral 3 14B 2512",+ "provider": "mistralai",+ "local_id": "ministral-14b-2512",+ "context_length": 262144,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.2,+ "output_price_per_mtok": 0.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "mistralai/ministral-8b-2512": {+ "id": "mistralai/ministral-8b-2512",+ "name": "Mistral: Ministral 3 8B 2512",+ "provider": "mistralai",+ "local_id": "ministral-8b-2512",+ "context_length": 262144,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.15,+ "output_price_per_mtok": 0.15,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "mistralai/ministral-3b-2512": {+ "id": "mistralai/ministral-3b-2512",+ "name": "Mistral: Ministral 3 3B 2512",+ "provider": "mistralai",+ "local_id": "ministral-3b-2512",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.1,+ "output_price_per_mtok": 0.1,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "mistralai/mistral-large-2512": {+ "id": "mistralai/mistral-large-2512",+ "name": "Mistral: Mistral Large 3 2512",+ "provider": "mistralai",+ "local_id": "mistral-large-2512",+ "context_length": 262144,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.5,+ "output_price_per_mtok": 1.5,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "arcee-ai/trinity-mini:free": {+ "id": "arcee-ai/trinity-mini:free",+ "name": "Arcee AI: Trinity Mini (free)",+ "provider": "arcee-ai",+ "local_id": "trinity-mini:free",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "arcee-ai/trinity-mini": {+ "id": "arcee-ai/trinity-mini",+ "name": "Arcee AI: Trinity Mini",+ "provider": "arcee-ai",+ "local_id": "trinity-mini",+ "context_length": 131072,+ "max_output_tokens": 131072,+ "input_price_per_mtok": 0.045,+ "output_price_per_mtok": 0.15,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "deepseek/deepseek-v3.2-speciale": {+ "id": "deepseek/deepseek-v3.2-speciale",+ "name": "DeepSeek: DeepSeek V3.2 Speciale",+ "provider": "deepseek",+ "local_id": "deepseek-v3.2-speciale",+ "context_length": 163840,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 0.4,+ "output_price_per_mtok": 1.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "temperature",+ "top_p"+ ]+ },+ "deepseek/deepseek-v3.2": {+ "id": "deepseek/deepseek-v3.2",+ "name": "DeepSeek: DeepSeek V3.2",+ "provider": "deepseek",+ "local_id": "deepseek-v3.2",+ "context_length": 163840,+ "max_output_tokens": 163840,+ "input_price_per_mtok": 0.25,+ "output_price_per_mtok": 0.4,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_logprobs",+ "top_p"+ ]+ },+ "prime-intellect/intellect-3": {+ "id": "prime-intellect/intellect-3",+ "name": "Prime Intellect: INTELLECT-3",+ "provider": "prime-intellect",+ "local_id": "intellect-3",+ "context_length": 131072,+ "max_output_tokens": 131072,+ "input_price_per_mtok": 0.2,+ "output_price_per_mtok": 1.1,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "anthropic/claude-opus-4.5": {+ "id": "anthropic/claude-opus-4.5",+ "name": "Anthropic: Claude Opus 4.5",+ "provider": "anthropic",+ "local_id": "claude-opus-4.5",+ "context_length": 200000,+ "max_output_tokens": 64000,+ "input_price_per_mtok": 5.0,+ "output_price_per_mtok": 25.0,+ "supports_thinking": true,+ "thinking_type": "budget",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "verbosity"+ ]+ },+ "allenai/olmo-3-32b-think": {+ "id": "allenai/olmo-3-32b-think",+ "name": "AllenAI: Olmo 3 32B Think",+ "provider": "allenai",+ "local_id": "olmo-3-32b-think",+ "context_length": 65536,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 0.15,+ "output_price_per_mtok": 0.5,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "allenai/olmo-3-7b-instruct": {+ "id": "allenai/olmo-3-7b-instruct",+ "name": "AllenAI: Olmo 3 7B Instruct",+ "provider": "allenai",+ "local_id": "olmo-3-7b-instruct",+ "context_length": 65536,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 0.1,+ "output_price_per_mtok": 0.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "allenai/olmo-3-7b-think": {+ "id": "allenai/olmo-3-7b-think",+ "name": "AllenAI: Olmo 3 7B Think",+ "provider": "allenai",+ "local_id": "olmo-3-7b-think",+ "context_length": 65536,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 0.12,+ "output_price_per_mtok": 0.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "gemini/gemini-3-pro-image-preview": {+ "id": "google/gemini-3-pro-image-preview",+ "name": "Google: Nano Banana Pro (Gemini 3 Pro Image Preview)",+ "provider": "gemini",+ "local_id": "gemini-3-pro-image-preview",+ "context_length": 65536,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 2.0,+ "output_price_per_mtok": 12.0,+ "supports_thinking": true,+ "thinking_type": "budget",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_p"+ ]+ },+ "xai/grok-4.1-fast": {+ "id": "x-ai/grok-4.1-fast",+ "name": "xAI: Grok 4.1 Fast",+ "provider": "xai",+ "local_id": "grok-4.1-fast",+ "context_length": 2000000,+ "max_output_tokens": 30000,+ "input_price_per_mtok": 0.2,+ "output_price_per_mtok": 0.5,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "logprobs",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p"+ ]+ },+ "gemini/gemini-3-pro-preview": {+ "id": "google/gemini-3-pro-preview",+ "name": "Google: Gemini 3 Pro Preview",+ "provider": "gemini",+ "local_id": "gemini-3-pro-preview",+ "context_length": 1048576,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 2.0,+ "output_price_per_mtok": 12.0,+ "supports_thinking": true,+ "thinking_type": "budget",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "deepcogito/cogito-v2.1-671b": {+ "id": "deepcogito/cogito-v2.1-671b",+ "name": "Deep Cogito: Cogito v2.1 671B",+ "provider": "deepcogito",+ "local_id": "cogito-v2.1-671b",+ "context_length": 128000,+ "max_output_tokens": 0,+ "input_price_per_mtok": 1.25,+ "output_price_per_mtok": 1.25,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "stop",+ "structured_outputs",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "openai/gpt-5.1": {+ "id": "openai/gpt-5.1",+ "name": "OpenAI: GPT-5.1",+ "provider": "openai",+ "local_id": "gpt-5.1",+ "context_length": 400000,+ "max_output_tokens": 128000,+ "input_price_per_mtok": 1.25,+ "output_price_per_mtok": 10.0,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "tool_choice",+ "tools"+ ]+ },+ "openai/gpt-5.1-chat": {+ "id": "openai/gpt-5.1-chat",+ "name": "OpenAI: GPT-5.1 Chat",+ "provider": "openai",+ "local_id": "gpt-5.1-chat",+ "context_length": 128000,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 1.25,+ "output_price_per_mtok": 10.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "response_format",+ "seed",+ "structured_outputs",+ "tool_choice",+ "tools"+ ]+ },+ "openai/gpt-5.1-codex": {+ "id": "openai/gpt-5.1-codex",+ "name": "OpenAI: GPT-5.1-Codex",+ "provider": "openai",+ "local_id": "gpt-5.1-codex",+ "context_length": 400000,+ "max_output_tokens": 128000,+ "input_price_per_mtok": 1.25,+ "output_price_per_mtok": 10.0,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "tool_choice",+ "tools"+ ]+ },+ "openai/gpt-5.1-codex-mini": {+ "id": "openai/gpt-5.1-codex-mini",+ "name": "OpenAI: GPT-5.1-Codex-Mini",+ "provider": "openai",+ "local_id": "gpt-5.1-codex-mini",+ "context_length": 400000,+ "max_output_tokens": 100000,+ "input_price_per_mtok": 0.25,+ "output_price_per_mtok": 2.0,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "tool_choice",+ "tools"+ ]+ },+ "kwaipilot/kat-coder-pro": {+ "id": "kwaipilot/kat-coder-pro",+ "name": "Kwaipilot: KAT-Coder-Pro V1",+ "provider": "kwaipilot",+ "local_id": "kat-coder-pro",+ "context_length": 256000,+ "max_output_tokens": 128000,+ "input_price_per_mtok": 0.207,+ "output_price_per_mtok": 0.828,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "moonshotai/kimi-k2-thinking": {+ "id": "moonshotai/kimi-k2-thinking",+ "name": "MoonshotAI: Kimi K2 Thinking",+ "provider": "moonshotai",+ "local_id": "kimi-k2-thinking",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.47,+ "output_price_per_mtok": 2.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "amazon/nova-premier-v1": {+ "id": "amazon/nova-premier-v1",+ "name": "Amazon: Nova Premier 1.0",+ "provider": "amazon",+ "local_id": "nova-premier-v1",+ "context_length": 1000000,+ "max_output_tokens": 32000,+ "input_price_per_mtok": 2.5,+ "output_price_per_mtok": 12.5,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "stop",+ "temperature",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "perplexity/sonar-pro-search": {+ "id": "perplexity/sonar-pro-search",+ "name": "Perplexity: Sonar Pro Search",+ "provider": "perplexity",+ "local_id": "sonar-pro-search",+ "context_length": 200000,+ "max_output_tokens": 8000,+ "input_price_per_mtok": 3.0,+ "output_price_per_mtok": 15.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "structured_outputs",+ "temperature",+ "top_k",+ "top_p",+ "web_search_options"+ ]+ },+ "mistralai/voxtral-small-24b-2507": {+ "id": "mistralai/voxtral-small-24b-2507",+ "name": "Mistral: Voxtral Small 24B 2507",+ "provider": "mistralai",+ "local_id": "voxtral-small-24b-2507",+ "context_length": 32000,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.1,+ "output_price_per_mtok": 0.3,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "openai/gpt-oss-safeguard-20b": {+ "id": "openai/gpt-oss-safeguard-20b",+ "name": "OpenAI: gpt-oss-safeguard-20b",+ "provider": "openai",+ "local_id": "gpt-oss-safeguard-20b",+ "context_length": 131072,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 0.075,+ "output_price_per_mtok": 0.3,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "nvidia/nemotron-nano-12b-v2-vl:free": {+ "id": "nvidia/nemotron-nano-12b-v2-vl:free",+ "name": "NVIDIA: Nemotron Nano 12B 2 VL (free)",+ "provider": "nvidia",+ "local_id": "nemotron-nano-12b-v2-vl:free",+ "context_length": 128000,+ "max_output_tokens": 128000,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "seed",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "nvidia/nemotron-nano-12b-v2-vl": {+ "id": "nvidia/nemotron-nano-12b-v2-vl",+ "name": "NVIDIA: Nemotron Nano 12B 2 VL",+ "provider": "nvidia",+ "local_id": "nemotron-nano-12b-v2-vl",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.2,+ "output_price_per_mtok": 0.6,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "minimax/minimax-m2": {+ "id": "minimax/minimax-m2",+ "name": "MiniMax: MiniMax M2",+ "provider": "minimax",+ "local_id": "minimax-m2",+ "context_length": 196608,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 0.255,+ "output_price_per_mtok": 1.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "qwen/qwen3-vl-32b-instruct": {+ "id": "qwen/qwen3-vl-32b-instruct",+ "name": "Qwen: Qwen3 VL 32B Instruct",+ "provider": "qwen",+ "local_id": "qwen3-vl-32b-instruct",+ "context_length": 131072,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.104,+ "output_price_per_mtok": 0.416,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "liquid/lfm2-8b-a1b": {+ "id": "liquid/lfm2-8b-a1b",+ "name": "LiquidAI: LFM2-8B-A1B",+ "provider": "liquid",+ "local_id": "lfm2-8b-a1b",+ "context_length": 32768,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.01,+ "output_price_per_mtok": 0.02,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "seed",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "liquid/lfm-2.2-6b": {+ "id": "liquid/lfm-2.2-6b",+ "name": "LiquidAI: LFM2-2.6B",+ "provider": "liquid",+ "local_id": "lfm-2.2-6b",+ "context_length": 32768,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.01,+ "output_price_per_mtok": 0.02,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "seed",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "ibm-granite/granite-4.0-h-micro": {+ "id": "ibm-granite/granite-4.0-h-micro",+ "name": "IBM: Granite 4.0 Micro",+ "provider": "ibm-granite",+ "local_id": "granite-4.0-h-micro",+ "context_length": 131000,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.017,+ "output_price_per_mtok": 0.11,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "repetition_penalty",+ "seed",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "openai/gpt-5-image-mini": {+ "id": "openai/gpt-5-image-mini",+ "name": "OpenAI: GPT-5 Image Mini",+ "provider": "openai",+ "local_id": "gpt-5-image-mini",+ "context_length": 400000,+ "max_output_tokens": 128000,+ "input_price_per_mtok": 2.5,+ "output_price_per_mtok": 2.0,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p"+ ]+ },+ "anthropic/claude-haiku-4.5": {+ "id": "anthropic/claude-haiku-4.5",+ "name": "Anthropic: Claude Haiku 4.5",+ "provider": "anthropic",+ "local_id": "claude-haiku-4.5",+ "context_length": 200000,+ "max_output_tokens": 64000,+ "input_price_per_mtok": 1.0,+ "output_price_per_mtok": 5.0,+ "supports_thinking": true,+ "thinking_type": "budget",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "qwen/qwen3-vl-8b-thinking": {+ "id": "qwen/qwen3-vl-8b-thinking",+ "name": "Qwen: Qwen3 VL 8B Thinking",+ "provider": "qwen",+ "local_id": "qwen3-vl-8b-thinking",+ "context_length": 131072,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.117,+ "output_price_per_mtok": 1.365,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "qwen/qwen3-vl-8b-instruct": {+ "id": "qwen/qwen3-vl-8b-instruct",+ "name": "Qwen: Qwen3 VL 8B Instruct",+ "provider": "qwen",+ "local_id": "qwen3-vl-8b-instruct",+ "context_length": 131072,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.08,+ "output_price_per_mtok": 0.5,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "openai/gpt-5-image": {+ "id": "openai/gpt-5-image",+ "name": "OpenAI: GPT-5 Image",+ "provider": "openai",+ "local_id": "gpt-5-image",+ "context_length": 400000,+ "max_output_tokens": 128000,+ "input_price_per_mtok": 10.0,+ "output_price_per_mtok": 10.0,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p"+ ]+ },+ "openai/o3-deep-research": {+ "id": "openai/o3-deep-research",+ "name": "OpenAI: o3 Deep Research",+ "provider": "openai",+ "local_id": "o3-deep-research",+ "context_length": 200000,+ "max_output_tokens": 100000,+ "input_price_per_mtok": 10.0,+ "output_price_per_mtok": 40.0,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 1.0,+ 1.0+ ],+ "is_reasoning_model": true,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p"+ ]+ },+ "openai/o4-mini-deep-research": {+ "id": "openai/o4-mini-deep-research",+ "name": "OpenAI: o4 Mini Deep Research",+ "provider": "openai",+ "local_id": "o4-mini-deep-research",+ "context_length": 200000,+ "max_output_tokens": 100000,+ "input_price_per_mtok": 2.0,+ "output_price_per_mtok": 8.0,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 1.0,+ 1.0+ ],+ "is_reasoning_model": true,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p"+ ]+ },+ "nvidia/llama-3.3-nemotron-super-49b-v1.5": {+ "id": "nvidia/llama-3.3-nemotron-super-49b-v1.5",+ "name": "NVIDIA: Llama 3.3 Nemotron Super 49B V1.5",+ "provider": "nvidia",+ "local_id": "llama-3.3-nemotron-super-49b-v1.5",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.1,+ "output_price_per_mtok": 0.4,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "baidu/ernie-4.5-21b-a3b-thinking": {+ "id": "baidu/ernie-4.5-21b-a3b-thinking",+ "name": "Baidu: ERNIE 4.5 21B A3B Thinking",+ "provider": "baidu",+ "local_id": "ernie-4.5-21b-a3b-thinking",+ "context_length": 131072,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 0.07,+ "output_price_per_mtok": 0.28,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "seed",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "gemini/gemini-2.5-flash-image": {+ "id": "google/gemini-2.5-flash-image",+ "name": "Google: Nano Banana (Gemini 2.5 Flash Image)",+ "provider": "gemini",+ "local_id": "gemini-2.5-flash-image",+ "context_length": 32768,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.3,+ "output_price_per_mtok": 2.5,+ "supports_thinking": true,+ "thinking_type": "budget",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_p"+ ]+ },+ "qwen/qwen3-vl-30b-a3b-thinking": {+ "id": "qwen/qwen3-vl-30b-a3b-thinking",+ "name": "Qwen: Qwen3 VL 30B A3B Thinking",+ "provider": "qwen",+ "local_id": "qwen3-vl-30b-a3b-thinking",+ "context_length": 131072,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "qwen/qwen3-vl-30b-a3b-instruct": {+ "id": "qwen/qwen3-vl-30b-a3b-instruct",+ "name": "Qwen: Qwen3 VL 30B A3B Instruct",+ "provider": "qwen",+ "local_id": "qwen3-vl-30b-a3b-instruct",+ "context_length": 131072,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.13,+ "output_price_per_mtok": 0.52,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "openai/gpt-5-pro": {+ "id": "openai/gpt-5-pro",+ "name": "OpenAI: GPT-5 Pro",+ "provider": "openai",+ "local_id": "gpt-5-pro",+ "context_length": 400000,+ "max_output_tokens": 128000,+ "input_price_per_mtok": 15.0,+ "output_price_per_mtok": 120.0,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "tool_choice",+ "tools"+ ]+ },+ "z-ai/glm-4.6": {+ "id": "z-ai/glm-4.6",+ "name": "Z.ai: GLM 4.6",+ "provider": "z-ai",+ "local_id": "glm-4.6",+ "context_length": 202752,+ "max_output_tokens": 131072,+ "input_price_per_mtok": 0.35,+ "output_price_per_mtok": 1.71,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "z-ai/glm-4.6:exacto": {+ "id": "z-ai/glm-4.6:exacto",+ "name": "Z.ai: GLM 4.6 (exacto)",+ "provider": "z-ai",+ "local_id": "glm-4.6:exacto",+ "context_length": 204800,+ "max_output_tokens": 131072,+ "input_price_per_mtok": 0.44,+ "output_price_per_mtok": 1.76,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "anthropic/claude-sonnet-4.5": {+ "id": "anthropic/claude-sonnet-4.5",+ "name": "Anthropic: Claude Sonnet 4.5",+ "provider": "anthropic",+ "local_id": "claude-sonnet-4.5",+ "context_length": 1000000,+ "max_output_tokens": 64000,+ "input_price_per_mtok": 3.0,+ "output_price_per_mtok": 15.0,+ "supports_thinking": true,+ "thinking_type": "budget",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "deepseek/deepseek-v3.2-exp": {+ "id": "deepseek/deepseek-v3.2-exp",+ "name": "DeepSeek: DeepSeek V3.2 Exp",+ "provider": "deepseek",+ "local_id": "deepseek-v3.2-exp",+ "context_length": 163840,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 0.27,+ "output_price_per_mtok": 0.41,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "thedrummer/cydonia-24b-v4.1": {+ "id": "thedrummer/cydonia-24b-v4.1",+ "name": "TheDrummer: Cydonia 24B V4.1",+ "provider": "thedrummer",+ "local_id": "cydonia-24b-v4.1",+ "context_length": 131072,+ "max_output_tokens": 131072,+ "input_price_per_mtok": 0.3,+ "output_price_per_mtok": 0.5,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "relace/relace-apply-3": {+ "id": "relace/relace-apply-3",+ "name": "Relace: Relace Apply 3",+ "provider": "relace",+ "local_id": "relace-apply-3",+ "context_length": 256000,+ "max_output_tokens": 128000,+ "input_price_per_mtok": 0.85,+ "output_price_per_mtok": 1.25,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "seed",+ "stop"+ ]+ },+ "gemini/gemini-2.5-flash-lite-preview-09-2025": {+ "id": "google/gemini-2.5-flash-lite-preview-09-2025",+ "name": "Google: Gemini 2.5 Flash Lite Preview 09-2025",+ "provider": "gemini",+ "local_id": "gemini-2.5-flash-lite-preview-09-2025",+ "context_length": 1048576,+ "max_output_tokens": 65535,+ "input_price_per_mtok": 0.1,+ "output_price_per_mtok": 0.4,+ "supports_thinking": true,+ "thinking_type": "budget",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "qwen/qwen3-vl-235b-a22b-thinking": {+ "id": "qwen/qwen3-vl-235b-a22b-thinking",+ "name": "Qwen: Qwen3 VL 235B A22B Thinking",+ "provider": "qwen",+ "local_id": "qwen3-vl-235b-a22b-thinking",+ "context_length": 131072,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "qwen/qwen3-vl-235b-a22b-instruct": {+ "id": "qwen/qwen3-vl-235b-a22b-instruct",+ "name": "Qwen: Qwen3 VL 235B A22B Instruct",+ "provider": "qwen",+ "local_id": "qwen3-vl-235b-a22b-instruct",+ "context_length": 262144,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.2,+ "output_price_per_mtok": 0.88,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "qwen/qwen3-max": {+ "id": "qwen/qwen3-max",+ "name": "Qwen: Qwen3 Max",+ "provider": "qwen",+ "local_id": "qwen3-max",+ "context_length": 262144,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 1.2,+ "output_price_per_mtok": 6.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "qwen/qwen3-coder-plus": {+ "id": "qwen/qwen3-coder-plus",+ "name": "Qwen: Qwen3 Coder Plus",+ "provider": "qwen",+ "local_id": "qwen3-coder-plus",+ "context_length": 1000000,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 1.0,+ "output_price_per_mtok": 5.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "openai/gpt-5-codex": {+ "id": "openai/gpt-5-codex",+ "name": "OpenAI: GPT-5 Codex",+ "provider": "openai",+ "local_id": "gpt-5-codex",+ "context_length": 400000,+ "max_output_tokens": 128000,+ "input_price_per_mtok": 1.25,+ "output_price_per_mtok": 10.0,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "tool_choice",+ "tools"+ ]+ },+ "deepseek/deepseek-v3.1-terminus:exacto": {+ "id": "deepseek/deepseek-v3.1-terminus:exacto",+ "name": "DeepSeek: DeepSeek V3.1 Terminus (exacto)",+ "provider": "deepseek",+ "local_id": "deepseek-v3.1-terminus:exacto",+ "context_length": 163840,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.21,+ "output_price_per_mtok": 0.79,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "deepseek/deepseek-v3.1-terminus": {+ "id": "deepseek/deepseek-v3.1-terminus",+ "name": "DeepSeek: DeepSeek V3.1 Terminus",+ "provider": "deepseek",+ "local_id": "deepseek-v3.1-terminus",+ "context_length": 163840,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.21,+ "output_price_per_mtok": 0.79,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "xai/grok-4-fast": {+ "id": "x-ai/grok-4-fast",+ "name": "xAI: Grok 4 Fast",+ "provider": "xai",+ "local_id": "grok-4-fast",+ "context_length": 2000000,+ "max_output_tokens": 30000,+ "input_price_per_mtok": 0.2,+ "output_price_per_mtok": 0.5,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "logprobs",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p"+ ]+ },+ "alibaba/tongyi-deepresearch-30b-a3b": {+ "id": "alibaba/tongyi-deepresearch-30b-a3b",+ "name": "Tongyi DeepResearch 30B A3B",+ "provider": "alibaba",+ "local_id": "tongyi-deepresearch-30b-a3b",+ "context_length": 131072,+ "max_output_tokens": 131072,+ "input_price_per_mtok": 0.09,+ "output_price_per_mtok": 0.45,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "qwen/qwen3-coder-flash": {+ "id": "qwen/qwen3-coder-flash",+ "name": "Qwen: Qwen3 Coder Flash",+ "provider": "qwen",+ "local_id": "qwen3-coder-flash",+ "context_length": 1000000,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 0.3,+ "output_price_per_mtok": 1.5,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "opengvlab/internvl3-78b": {+ "id": "opengvlab/internvl3-78b",+ "name": "OpenGVLab: InternVL3 78B",+ "provider": "opengvlab",+ "local_id": "internvl3-78b",+ "context_length": 32768,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.15,+ "output_price_per_mtok": 0.6,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "qwen/qwen3-next-80b-a3b-thinking": {+ "id": "qwen/qwen3-next-80b-a3b-thinking",+ "name": "Qwen: Qwen3 Next 80B A3B Thinking",+ "provider": "qwen",+ "local_id": "qwen3-next-80b-a3b-thinking",+ "context_length": 128000,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.15,+ "output_price_per_mtok": 1.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "qwen/qwen3-next-80b-a3b-instruct:free": {+ "id": "qwen/qwen3-next-80b-a3b-instruct:free",+ "name": "Qwen: Qwen3 Next 80B A3B Instruct (free)",+ "provider": "qwen",+ "local_id": "qwen3-next-80b-a3b-instruct:free",+ "context_length": 262144,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "qwen/qwen3-next-80b-a3b-instruct": {+ "id": "qwen/qwen3-next-80b-a3b-instruct",+ "name": "Qwen: Qwen3 Next 80B A3B Instruct",+ "provider": "qwen",+ "local_id": "qwen3-next-80b-a3b-instruct",+ "context_length": 262144,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.09,+ "output_price_per_mtok": 1.1,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "meituan/longcat-flash-chat": {+ "id": "meituan/longcat-flash-chat",+ "name": "Meituan: LongCat Flash Chat",+ "provider": "meituan",+ "local_id": "longcat-flash-chat",+ "context_length": 131072,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.2,+ "output_price_per_mtok": 0.8,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "response_format",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "qwen/qwen-plus-2025-07-28:thinking": {+ "id": "qwen/qwen-plus-2025-07-28:thinking",+ "name": "Qwen: Qwen Plus 0728 (thinking)",+ "provider": "qwen",+ "local_id": "qwen-plus-2025-07-28:thinking",+ "context_length": 1000000,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.4,+ "output_price_per_mtok": 1.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "qwen/qwen-plus-2025-07-28": {+ "id": "qwen/qwen-plus-2025-07-28",+ "name": "Qwen: Qwen Plus 0728",+ "provider": "qwen",+ "local_id": "qwen-plus-2025-07-28",+ "context_length": 1000000,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.4,+ "output_price_per_mtok": 1.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "nvidia/nemotron-nano-9b-v2:free": {+ "id": "nvidia/nemotron-nano-9b-v2:free",+ "name": "NVIDIA: Nemotron Nano 9B V2 (free)",+ "provider": "nvidia",+ "local_id": "nemotron-nano-9b-v2:free",+ "context_length": 128000,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "nvidia/nemotron-nano-9b-v2": {+ "id": "nvidia/nemotron-nano-9b-v2",+ "name": "NVIDIA: Nemotron Nano 9B V2",+ "provider": "nvidia",+ "local_id": "nemotron-nano-9b-v2",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.04,+ "output_price_per_mtok": 0.16,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "moonshotai/kimi-k2-0905": {+ "id": "moonshotai/kimi-k2-0905",+ "name": "MoonshotAI: Kimi K2 0905",+ "provider": "moonshotai",+ "local_id": "kimi-k2-0905",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.4,+ "output_price_per_mtok": 2.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_logprobs",+ "top_p"+ ]+ },+ "moonshotai/kimi-k2-0905:exacto": {+ "id": "moonshotai/kimi-k2-0905:exacto",+ "name": "MoonshotAI: Kimi K2 0905 (exacto)",+ "provider": "moonshotai",+ "local_id": "kimi-k2-0905:exacto",+ "context_length": 262144,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.6,+ "output_price_per_mtok": 2.5,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "qwen/qwen3-30b-a3b-thinking-2507": {+ "id": "qwen/qwen3-30b-a3b-thinking-2507",+ "name": "Qwen: Qwen3 30B A3B Thinking 2507",+ "provider": "qwen",+ "local_id": "qwen3-30b-a3b-thinking-2507",+ "context_length": 32768,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.051,+ "output_price_per_mtok": 0.34,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "xai/grok-code-fast-1": {+ "id": "x-ai/grok-code-fast-1",+ "name": "xAI: Grok Code Fast 1",+ "provider": "xai",+ "local_id": "grok-code-fast-1",+ "context_length": 256000,+ "max_output_tokens": 10000,+ "input_price_per_mtok": 0.2,+ "output_price_per_mtok": 1.5,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "logprobs",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p"+ ]+ },+ "nousresearch/hermes-4-70b": {+ "id": "nousresearch/hermes-4-70b",+ "name": "Nous: Hermes 4 70B",+ "provider": "nousresearch",+ "local_id": "hermes-4-70b",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.13,+ "output_price_per_mtok": 0.4,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "nousresearch/hermes-4-405b": {+ "id": "nousresearch/hermes-4-405b",+ "name": "Nous: Hermes 4 405B",+ "provider": "nousresearch",+ "local_id": "hermes-4-405b",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 1.0,+ "output_price_per_mtok": 3.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "deepseek/deepseek-chat-v3.1": {+ "id": "deepseek/deepseek-chat-v3.1",+ "name": "DeepSeek: DeepSeek V3.1",+ "provider": "deepseek",+ "local_id": "deepseek-chat-v3.1",+ "context_length": 32768,+ "max_output_tokens": 7168,+ "input_price_per_mtok": 0.15,+ "output_price_per_mtok": 0.75,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_logprobs",+ "top_p"+ ]+ },+ "openai/gpt-4o-audio-preview": {+ "id": "openai/gpt-4o-audio-preview",+ "name": "OpenAI: GPT-4o Audio",+ "provider": "openai",+ "local_id": "gpt-4o-audio-preview",+ "context_length": 128000,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 2.5,+ "output_price_per_mtok": 10.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p"+ ]+ },+ "mistralai/mistral-medium-3.1": {+ "id": "mistralai/mistral-medium-3.1",+ "name": "Mistral: Mistral Medium 3.1",+ "provider": "mistralai",+ "local_id": "mistral-medium-3.1",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.4,+ "output_price_per_mtok": 2.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "baidu/ernie-4.5-21b-a3b": {+ "id": "baidu/ernie-4.5-21b-a3b",+ "name": "Baidu: ERNIE 4.5 21B A3B",+ "provider": "baidu",+ "local_id": "ernie-4.5-21b-a3b",+ "context_length": 120000,+ "max_output_tokens": 8000,+ "input_price_per_mtok": 0.07,+ "output_price_per_mtok": 0.28,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "repetition_penalty",+ "seed",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "baidu/ernie-4.5-vl-28b-a3b": {+ "id": "baidu/ernie-4.5-vl-28b-a3b",+ "name": "Baidu: ERNIE 4.5 VL 28B A3B",+ "provider": "baidu",+ "local_id": "ernie-4.5-vl-28b-a3b",+ "context_length": 30000,+ "max_output_tokens": 8000,+ "input_price_per_mtok": 0.14,+ "output_price_per_mtok": 0.56,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "seed",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "z-ai/glm-4.5v": {+ "id": "z-ai/glm-4.5v",+ "name": "Z.ai: GLM 4.5V",+ "provider": "z-ai",+ "local_id": "glm-4.5v",+ "context_length": 65536,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 0.6,+ "output_price_per_mtok": 1.8,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "ai21/jamba-large-1.7": {+ "id": "ai21/jamba-large-1.7",+ "name": "AI21: Jamba Large 1.7",+ "provider": "ai21",+ "local_id": "jamba-large-1.7",+ "context_length": 256000,+ "max_output_tokens": 4096,+ "input_price_per_mtok": 2.0,+ "output_price_per_mtok": 8.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "response_format",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "openai/gpt-5-chat": {+ "id": "openai/gpt-5-chat",+ "name": "OpenAI: GPT-5 Chat",+ "provider": "openai",+ "local_id": "gpt-5-chat",+ "context_length": 128000,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 1.25,+ "output_price_per_mtok": 10.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "response_format",+ "seed",+ "structured_outputs"+ ]+ },+ "openai/gpt-5": {+ "id": "openai/gpt-5",+ "name": "OpenAI: GPT-5",+ "provider": "openai",+ "local_id": "gpt-5",+ "context_length": 400000,+ "max_output_tokens": 128000,+ "input_price_per_mtok": 1.25,+ "output_price_per_mtok": 10.0,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "tool_choice",+ "tools"+ ]+ },+ "openai/gpt-5-mini": {+ "id": "openai/gpt-5-mini",+ "name": "OpenAI: GPT-5 Mini",+ "provider": "openai",+ "local_id": "gpt-5-mini",+ "context_length": 400000,+ "max_output_tokens": 128000,+ "input_price_per_mtok": 0.25,+ "output_price_per_mtok": 2.0,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "tool_choice",+ "tools"+ ]+ },+ "openai/gpt-5-nano": {+ "id": "openai/gpt-5-nano",+ "name": "OpenAI: GPT-5 Nano",+ "provider": "openai",+ "local_id": "gpt-5-nano",+ "context_length": 400000,+ "max_output_tokens": 128000,+ "input_price_per_mtok": 0.05,+ "output_price_per_mtok": 0.4,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "tool_choice",+ "tools"+ ]+ },+ "openai/gpt-oss-120b:free": {+ "id": "openai/gpt-oss-120b:free",+ "name": "OpenAI: gpt-oss-120b (free)",+ "provider": "openai",+ "local_id": "gpt-oss-120b:free",+ "context_length": 131072,+ "max_output_tokens": 131072,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "seed",+ "stop",+ "temperature",+ "tool_choice",+ "tools"+ ]+ },+ "openai/gpt-oss-120b": {+ "id": "openai/gpt-oss-120b",+ "name": "OpenAI: gpt-oss-120b",+ "provider": "openai",+ "local_id": "gpt-oss-120b",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.039,+ "output_price_per_mtok": 0.19,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "reasoning_effort",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_logprobs",+ "top_p"+ ]+ },+ "openai/gpt-oss-120b:exacto": {+ "id": "openai/gpt-oss-120b:exacto",+ "name": "OpenAI: gpt-oss-120b (exacto)",+ "provider": "openai",+ "local_id": "gpt-oss-120b:exacto",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.039,+ "output_price_per_mtok": 0.19,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "openai/gpt-oss-20b:free": {+ "id": "openai/gpt-oss-20b:free",+ "name": "OpenAI: gpt-oss-20b (free)",+ "provider": "openai",+ "local_id": "gpt-oss-20b:free",+ "context_length": 131072,+ "max_output_tokens": 131072,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "seed",+ "stop",+ "temperature",+ "tool_choice",+ "tools"+ ]+ },+ "openai/gpt-oss-20b": {+ "id": "openai/gpt-oss-20b",+ "name": "OpenAI: gpt-oss-20b",+ "provider": "openai",+ "local_id": "gpt-oss-20b",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.03,+ "output_price_per_mtok": 0.14,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "reasoning_effort",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "anthropic/claude-opus-4.1": {+ "id": "anthropic/claude-opus-4.1",+ "name": "Anthropic: Claude Opus 4.1",+ "provider": "anthropic",+ "local_id": "claude-opus-4.1",+ "context_length": 200000,+ "max_output_tokens": 32000,+ "input_price_per_mtok": 15.0,+ "output_price_per_mtok": 75.0,+ "supports_thinking": true,+ "thinking_type": "budget",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "mistralai/codestral-2508": {+ "id": "mistralai/codestral-2508",+ "name": "Mistral: Codestral 2508",+ "provider": "mistralai",+ "local_id": "codestral-2508",+ "context_length": 256000,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.3,+ "output_price_per_mtok": 0.9,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "qwen/qwen3-coder-30b-a3b-instruct": {+ "id": "qwen/qwen3-coder-30b-a3b-instruct",+ "name": "Qwen: Qwen3 Coder 30B A3B Instruct",+ "provider": "qwen",+ "local_id": "qwen3-coder-30b-a3b-instruct",+ "context_length": 160000,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.07,+ "output_price_per_mtok": 0.27,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "qwen/qwen3-30b-a3b-instruct-2507": {+ "id": "qwen/qwen3-30b-a3b-instruct-2507",+ "name": "Qwen: Qwen3 30B A3B Instruct 2507",+ "provider": "qwen",+ "local_id": "qwen3-30b-a3b-instruct-2507",+ "context_length": 262144,+ "max_output_tokens": 262144,+ "input_price_per_mtok": 0.09,+ "output_price_per_mtok": 0.3,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "z-ai/glm-4.5": {+ "id": "z-ai/glm-4.5",+ "name": "Z.ai: GLM 4.5",+ "provider": "z-ai",+ "local_id": "glm-4.5",+ "context_length": 131000,+ "max_output_tokens": 131000,+ "input_price_per_mtok": 0.55,+ "output_price_per_mtok": 2.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "z-ai/glm-4.5-air:free": {+ "id": "z-ai/glm-4.5-air:free",+ "name": "Z.ai: GLM 4.5 Air (free)",+ "provider": "z-ai",+ "local_id": "glm-4.5-air:free",+ "context_length": 131072,+ "max_output_tokens": 96000,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "z-ai/glm-4.5-air": {+ "id": "z-ai/glm-4.5-air",+ "name": "Z.ai: GLM 4.5 Air",+ "provider": "z-ai",+ "local_id": "glm-4.5-air",+ "context_length": 131072,+ "max_output_tokens": 98304,+ "input_price_per_mtok": 0.13,+ "output_price_per_mtok": 0.85,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "qwen/qwen3-235b-a22b-thinking-2507": {+ "id": "qwen/qwen3-235b-a22b-thinking-2507",+ "name": "Qwen: Qwen3 235B A22B Thinking 2507",+ "provider": "qwen",+ "local_id": "qwen3-235b-a22b-thinking-2507",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "z-ai/glm-4-32b": {+ "id": "z-ai/glm-4-32b",+ "name": "Z.ai: GLM 4 32B ",+ "provider": "z-ai",+ "local_id": "glm-4-32b",+ "context_length": 128000,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.1,+ "output_price_per_mtok": 0.1,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "qwen/qwen3-coder:free": {+ "id": "qwen/qwen3-coder:free",+ "name": "Qwen: Qwen3 Coder 480B A35B (free)",+ "provider": "qwen",+ "local_id": "qwen3-coder:free",+ "context_length": 262000,+ "max_output_tokens": 262000,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "qwen/qwen3-coder": {+ "id": "qwen/qwen3-coder",+ "name": "Qwen: Qwen3 Coder 480B A35B",+ "provider": "qwen",+ "local_id": "qwen3-coder",+ "context_length": 262144,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.22,+ "output_price_per_mtok": 1.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "qwen/qwen3-coder:exacto": {+ "id": "qwen/qwen3-coder:exacto",+ "name": "Qwen: Qwen3 Coder 480B A35B (exacto)",+ "provider": "qwen",+ "local_id": "qwen3-coder:exacto",+ "context_length": 262144,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 0.22,+ "output_price_per_mtok": 1.8,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "bytedance/ui-tars-1.5-7b": {+ "id": "bytedance/ui-tars-1.5-7b",+ "name": "ByteDance: UI-TARS 7B ",+ "provider": "bytedance",+ "local_id": "ui-tars-1.5-7b",+ "context_length": 128000,+ "max_output_tokens": 2048,+ "input_price_per_mtok": 0.1,+ "output_price_per_mtok": 0.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "presence_penalty",+ "repetition_penalty",+ "seed",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "gemini/gemini-2.5-flash-lite": {+ "id": "google/gemini-2.5-flash-lite",+ "name": "Google: Gemini 2.5 Flash Lite",+ "provider": "gemini",+ "local_id": "gemini-2.5-flash-lite",+ "context_length": 1048576,+ "max_output_tokens": 65535,+ "input_price_per_mtok": 0.1,+ "output_price_per_mtok": 0.4,+ "supports_thinking": true,+ "thinking_type": "budget",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "qwen/qwen3-235b-a22b-2507": {+ "id": "qwen/qwen3-235b-a22b-2507",+ "name": "Qwen: Qwen3 235B A22B Instruct 2507",+ "provider": "qwen",+ "local_id": "qwen3-235b-a22b-2507",+ "context_length": 262144,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.071,+ "output_price_per_mtok": 0.1,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "reasoning_effort",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_logprobs",+ "top_p"+ ]+ },+ "switchpoint/router": {+ "id": "switchpoint/router",+ "name": "Switchpoint Router",+ "provider": "switchpoint",+ "local_id": "router",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.85,+ "output_price_per_mtok": 3.4,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "seed",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "moonshotai/kimi-k2": {+ "id": "moonshotai/kimi-k2",+ "name": "MoonshotAI: Kimi K2 0711",+ "provider": "moonshotai",+ "local_id": "kimi-k2",+ "context_length": 131000,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.55,+ "output_price_per_mtok": 2.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logprobs",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_logprobs",+ "top_p"+ ]+ },+ "mistralai/devstral-medium": {+ "id": "mistralai/devstral-medium",+ "name": "Mistral: Devstral Medium",+ "provider": "mistralai",+ "local_id": "devstral-medium",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.4,+ "output_price_per_mtok": 2.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "mistralai/devstral-small": {+ "id": "mistralai/devstral-small",+ "name": "Mistral: Devstral Small 1.1",+ "provider": "mistralai",+ "local_id": "devstral-small",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.1,+ "output_price_per_mtok": 0.3,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "cognitivecomputations/dolphin-mistral-24b-venice-edition:free": {+ "id": "cognitivecomputations/dolphin-mistral-24b-venice-edition:free",+ "name": "Venice: Uncensored (free)",+ "provider": "cognitivecomputations",+ "local_id": "dolphin-mistral-24b-venice-edition:free",+ "context_length": 32768,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "stop",+ "structured_outputs",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "xai/grok-4": {+ "id": "x-ai/grok-4",+ "name": "xAI: Grok 4",+ "provider": "xai",+ "local_id": "grok-4",+ "context_length": 256000,+ "max_output_tokens": 0,+ "input_price_per_mtok": 3.0,+ "output_price_per_mtok": 15.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "logprobs",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p"+ ]+ },+ "gemini/gemma-3n-e2b-it:free": {+ "id": "google/gemma-3n-e2b-it:free",+ "name": "Google: Gemma 3n 2B (free)",+ "provider": "gemini",+ "local_id": "gemma-3n-e2b-it:free",+ "context_length": 8192,+ "max_output_tokens": 2048,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "top_p"+ ]+ },+ "tencent/hunyuan-a13b-instruct": {+ "id": "tencent/hunyuan-a13b-instruct",+ "name": "Tencent: Hunyuan A13B Instruct",+ "provider": "tencent",+ "local_id": "hunyuan-a13b-instruct",+ "context_length": 131072,+ "max_output_tokens": 131072,+ "input_price_per_mtok": 0.14,+ "output_price_per_mtok": 0.57,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "reasoning",+ "response_format",+ "structured_outputs",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "tngtech/deepseek-r1t2-chimera": {+ "id": "tngtech/deepseek-r1t2-chimera",+ "name": "TNG: DeepSeek R1T2 Chimera",+ "provider": "tngtech",+ "local_id": "deepseek-r1t2-chimera",+ "context_length": 163840,+ "max_output_tokens": 163840,+ "input_price_per_mtok": 0.25,+ "output_price_per_mtok": 0.85,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "morph/morph-v3-large": {+ "id": "morph/morph-v3-large",+ "name": "Morph: Morph V3 Large",+ "provider": "morph",+ "local_id": "morph-v3-large",+ "context_length": 262144,+ "max_output_tokens": 131072,+ "input_price_per_mtok": 0.9,+ "output_price_per_mtok": 1.9,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "stop",+ "temperature"+ ]+ },+ "morph/morph-v3-fast": {+ "id": "morph/morph-v3-fast",+ "name": "Morph: Morph V3 Fast",+ "provider": "morph",+ "local_id": "morph-v3-fast",+ "context_length": 81920,+ "max_output_tokens": 38000,+ "input_price_per_mtok": 0.8,+ "output_price_per_mtok": 1.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "stop",+ "temperature"+ ]+ },+ "baidu/ernie-4.5-vl-424b-a47b": {+ "id": "baidu/ernie-4.5-vl-424b-a47b",+ "name": "Baidu: ERNIE 4.5 VL 424B A47B ",+ "provider": "baidu",+ "local_id": "ernie-4.5-vl-424b-a47b",+ "context_length": 123000,+ "max_output_tokens": 16000,+ "input_price_per_mtok": 0.42,+ "output_price_per_mtok": 1.25,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "seed",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "baidu/ernie-4.5-300b-a47b": {+ "id": "baidu/ernie-4.5-300b-a47b",+ "name": "Baidu: ERNIE 4.5 300B A47B ",+ "provider": "baidu",+ "local_id": "ernie-4.5-300b-a47b",+ "context_length": 123000,+ "max_output_tokens": 12000,+ "input_price_per_mtok": 0.28,+ "output_price_per_mtok": 1.1,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "inception/mercury": {+ "id": "inception/mercury",+ "name": "Inception: Mercury",+ "provider": "inception",+ "local_id": "mercury",+ "context_length": 128000,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 0.25,+ "output_price_per_mtok": 1.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "mistralai/mistral-small-3.2-24b-instruct": {+ "id": "mistralai/mistral-small-3.2-24b-instruct",+ "name": "Mistral: Mistral Small 3.2 24B",+ "provider": "mistralai",+ "local_id": "mistral-small-3.2-24b-instruct",+ "context_length": 131072,+ "max_output_tokens": 131072,+ "input_price_per_mtok": 0.06,+ "output_price_per_mtok": 0.18,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "minimax/minimax-m1": {+ "id": "minimax/minimax-m1",+ "name": "MiniMax: MiniMax M1",+ "provider": "minimax",+ "local_id": "minimax-m1",+ "context_length": 1000000,+ "max_output_tokens": 40000,+ "input_price_per_mtok": 0.4,+ "output_price_per_mtok": 2.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "seed",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "gemini/gemini-2.5-flash": {+ "id": "google/gemini-2.5-flash",+ "name": "Google: Gemini 2.5 Flash",+ "provider": "gemini",+ "local_id": "gemini-2.5-flash",+ "context_length": 1048576,+ "max_output_tokens": 65535,+ "input_price_per_mtok": 0.3,+ "output_price_per_mtok": 2.5,+ "supports_thinking": true,+ "thinking_type": "budget",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "gemini/gemini-2.5-pro": {+ "id": "google/gemini-2.5-pro",+ "name": "Google: Gemini 2.5 Pro",+ "provider": "gemini",+ "local_id": "gemini-2.5-pro",+ "context_length": 1048576,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 1.25,+ "output_price_per_mtok": 10.0,+ "supports_thinking": true,+ "thinking_type": "budget",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "openai/o3-pro": {+ "id": "openai/o3-pro",+ "name": "OpenAI: o3 Pro",+ "provider": "openai",+ "local_id": "o3-pro",+ "context_length": 200000,+ "max_output_tokens": 100000,+ "input_price_per_mtok": 20.0,+ "output_price_per_mtok": 80.0,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 1.0,+ 1.0+ ],+ "is_reasoning_model": true,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "tool_choice",+ "tools"+ ]+ },+ "xai/grok-3-mini": {+ "id": "x-ai/grok-3-mini",+ "name": "xAI: Grok 3 Mini",+ "provider": "xai",+ "local_id": "grok-3-mini",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.3,+ "output_price_per_mtok": 0.5,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "logprobs",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p"+ ]+ },+ "xai/grok-3": {+ "id": "x-ai/grok-3",+ "name": "xAI: Grok 3",+ "provider": "xai",+ "local_id": "grok-3",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 3.0,+ "output_price_per_mtok": 15.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p"+ ]+ },+ "gemini/gemini-2.5-pro-preview": {+ "id": "google/gemini-2.5-pro-preview",+ "name": "Google: Gemini 2.5 Pro Preview 06-05",+ "provider": "gemini",+ "local_id": "gemini-2.5-pro-preview",+ "context_length": 1048576,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 1.25,+ "output_price_per_mtok": 10.0,+ "supports_thinking": true,+ "thinking_type": "budget",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "deepseek/deepseek-r1-0528": {+ "id": "deepseek/deepseek-r1-0528",+ "name": "DeepSeek: R1 0528",+ "provider": "deepseek",+ "local_id": "deepseek-r1-0528",+ "context_length": 163840,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 0.4,+ "output_price_per_mtok": 1.75,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_logprobs",+ "top_p"+ ]+ },+ "anthropic/claude-opus-4": {+ "id": "anthropic/claude-opus-4",+ "name": "Anthropic: Claude Opus 4",+ "provider": "anthropic",+ "local_id": "claude-opus-4",+ "context_length": 200000,+ "max_output_tokens": 32000,+ "input_price_per_mtok": 15.0,+ "output_price_per_mtok": 75.0,+ "supports_thinking": true,+ "thinking_type": "budget",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "anthropic/claude-sonnet-4": {+ "id": "anthropic/claude-sonnet-4",+ "name": "Anthropic: Claude Sonnet 4",+ "provider": "anthropic",+ "local_id": "claude-sonnet-4",+ "context_length": 1000000,+ "max_output_tokens": 64000,+ "input_price_per_mtok": 3.0,+ "output_price_per_mtok": 15.0,+ "supports_thinking": true,+ "thinking_type": "budget",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "gemini/gemma-3n-e4b-it:free": {+ "id": "google/gemma-3n-e4b-it:free",+ "name": "Google: Gemma 3n 4B (free)",+ "provider": "gemini",+ "local_id": "gemma-3n-e4b-it:free",+ "context_length": 8192,+ "max_output_tokens": 2048,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "top_p"+ ]+ },+ "gemini/gemma-3n-e4b-it": {+ "id": "google/gemma-3n-e4b-it",+ "name": "Google: Gemma 3n 4B",+ "provider": "gemini",+ "local_id": "gemma-3n-e4b-it",+ "context_length": 32768,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.02,+ "output_price_per_mtok": 0.04,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "mistralai/mistral-medium-3": {+ "id": "mistralai/mistral-medium-3",+ "name": "Mistral: Mistral Medium 3",+ "provider": "mistralai",+ "local_id": "mistral-medium-3",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.4,+ "output_price_per_mtok": 2.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "gemini/gemini-2.5-pro-preview-05-06": {+ "id": "google/gemini-2.5-pro-preview-05-06",+ "name": "Google: Gemini 2.5 Pro Preview 05-06",+ "provider": "gemini",+ "local_id": "gemini-2.5-pro-preview-05-06",+ "context_length": 1048576,+ "max_output_tokens": 65535,+ "input_price_per_mtok": 1.25,+ "output_price_per_mtok": 10.0,+ "supports_thinking": true,+ "thinking_type": "budget",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "arcee-ai/spotlight": {+ "id": "arcee-ai/spotlight",+ "name": "Arcee AI: Spotlight",+ "provider": "arcee-ai",+ "local_id": "spotlight",+ "context_length": 131072,+ "max_output_tokens": 65537,+ "input_price_per_mtok": 0.18,+ "output_price_per_mtok": 0.18,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "arcee-ai/maestro-reasoning": {+ "id": "arcee-ai/maestro-reasoning",+ "name": "Arcee AI: Maestro Reasoning",+ "provider": "arcee-ai",+ "local_id": "maestro-reasoning",+ "context_length": 131072,+ "max_output_tokens": 32000,+ "input_price_per_mtok": 0.9,+ "output_price_per_mtok": 3.3,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "arcee-ai/virtuoso-large": {+ "id": "arcee-ai/virtuoso-large",+ "name": "Arcee AI: Virtuoso Large",+ "provider": "arcee-ai",+ "local_id": "virtuoso-large",+ "context_length": 131072,+ "max_output_tokens": 64000,+ "input_price_per_mtok": 0.75,+ "output_price_per_mtok": 1.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "arcee-ai/coder-large": {+ "id": "arcee-ai/coder-large",+ "name": "Arcee AI: Coder Large",+ "provider": "arcee-ai",+ "local_id": "coder-large",+ "context_length": 32768,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.5,+ "output_price_per_mtok": 0.8,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "inception/mercury-coder": {+ "id": "inception/mercury-coder",+ "name": "Inception: Mercury Coder",+ "provider": "inception",+ "local_id": "mercury-coder",+ "context_length": 128000,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 0.25,+ "output_price_per_mtok": 1.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "qwen/qwen3-4b:free": {+ "id": "qwen/qwen3-4b:free",+ "name": "Qwen: Qwen3 4B (free)",+ "provider": "qwen",+ "local_id": "qwen3-4b:free",+ "context_length": 40960,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "response_format",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "meta/llama-guard-4-12b": {+ "id": "meta-llama/llama-guard-4-12b",+ "name": "Meta: Llama Guard 4 12B",+ "provider": "meta",+ "local_id": "llama-guard-4-12b",+ "context_length": 163840,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.18,+ "output_price_per_mtok": 0.18,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "qwen/qwen3-30b-a3b": {+ "id": "qwen/qwen3-30b-a3b",+ "name": "Qwen: Qwen3 30B A3B",+ "provider": "qwen",+ "local_id": "qwen3-30b-a3b",+ "context_length": 40960,+ "max_output_tokens": 40960,+ "input_price_per_mtok": 0.08,+ "output_price_per_mtok": 0.28,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "qwen/qwen3-8b": {+ "id": "qwen/qwen3-8b",+ "name": "Qwen: Qwen3 8B",+ "provider": "qwen",+ "local_id": "qwen3-8b",+ "context_length": 32000,+ "max_output_tokens": 8192,+ "input_price_per_mtok": 0.05,+ "output_price_per_mtok": 0.4,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "qwen/qwen3-14b": {+ "id": "qwen/qwen3-14b",+ "name": "Qwen: Qwen3 14B",+ "provider": "qwen",+ "local_id": "qwen3-14b",+ "context_length": 40960,+ "max_output_tokens": 40960,+ "input_price_per_mtok": 0.06,+ "output_price_per_mtok": 0.24,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "qwen/qwen3-32b": {+ "id": "qwen/qwen3-32b",+ "name": "Qwen: Qwen3 32B",+ "provider": "qwen",+ "local_id": "qwen3-32b",+ "context_length": 40960,+ "max_output_tokens": 40960,+ "input_price_per_mtok": 0.08,+ "output_price_per_mtok": 0.24,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "qwen/qwen3-235b-a22b": {+ "id": "qwen/qwen3-235b-a22b",+ "name": "Qwen: Qwen3 235B A22B",+ "provider": "qwen",+ "local_id": "qwen3-235b-a22b",+ "context_length": 131072,+ "max_output_tokens": 8192,+ "input_price_per_mtok": 0.455,+ "output_price_per_mtok": 1.82,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "response_format",+ "seed",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "openai/o4-mini-high": {+ "id": "openai/o4-mini-high",+ "name": "OpenAI: o4 Mini High",+ "provider": "openai",+ "local_id": "o4-mini-high",+ "context_length": 200000,+ "max_output_tokens": 100000,+ "input_price_per_mtok": 1.1,+ "output_price_per_mtok": 4.4,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 1.0,+ 1.0+ ],+ "is_reasoning_model": true,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "tool_choice",+ "tools"+ ]+ },+ "openai/o3": {+ "id": "openai/o3",+ "name": "OpenAI: o3",+ "provider": "openai",+ "local_id": "o3",+ "context_length": 200000,+ "max_output_tokens": 100000,+ "input_price_per_mtok": 2.0,+ "output_price_per_mtok": 8.0,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 1.0,+ 1.0+ ],+ "is_reasoning_model": true,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "tool_choice",+ "tools"+ ]+ },+ "openai/o4-mini": {+ "id": "openai/o4-mini",+ "name": "OpenAI: o4 Mini",+ "provider": "openai",+ "local_id": "o4-mini",+ "context_length": 200000,+ "max_output_tokens": 100000,+ "input_price_per_mtok": 1.1,+ "output_price_per_mtok": 4.4,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 1.0,+ 1.0+ ],+ "is_reasoning_model": true,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs",+ "tool_choice",+ "tools"+ ]+ },+ "qwen/qwen2.5-coder-7b-instruct": {+ "id": "qwen/qwen2.5-coder-7b-instruct",+ "name": "Qwen: Qwen2.5 Coder 7B Instruct",+ "provider": "qwen",+ "local_id": "qwen2.5-coder-7b-instruct",+ "context_length": 32768,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.03,+ "output_price_per_mtok": 0.09,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "structured_outputs",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "openai/gpt-4.1": {+ "id": "openai/gpt-4.1",+ "name": "OpenAI: GPT-4.1",+ "provider": "openai",+ "local_id": "gpt-4.1",+ "context_length": 1047576,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 2.0,+ "output_price_per_mtok": 8.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "response_format",+ "seed",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "openai/gpt-4.1-mini": {+ "id": "openai/gpt-4.1-mini",+ "name": "OpenAI: GPT-4.1 Mini",+ "provider": "openai",+ "local_id": "gpt-4.1-mini",+ "context_length": 1047576,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.4,+ "output_price_per_mtok": 1.6,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "response_format",+ "seed",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "openai/gpt-4.1-nano": {+ "id": "openai/gpt-4.1-nano",+ "name": "OpenAI: GPT-4.1 Nano",+ "provider": "openai",+ "local_id": "gpt-4.1-nano",+ "context_length": 1047576,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.1,+ "output_price_per_mtok": 0.4,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "response_format",+ "seed",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "eleutherai/llemma_7b": {+ "id": "eleutherai/llemma_7b",+ "name": "EleutherAI: Llemma 7b",+ "provider": "eleutherai",+ "local_id": "llemma_7b",+ "context_length": 4096,+ "max_output_tokens": 4096,+ "input_price_per_mtok": 0.8,+ "output_price_per_mtok": 1.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "seed",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "alfredpros/codellama-7b-instruct-solidity": {+ "id": "alfredpros/codellama-7b-instruct-solidity",+ "name": "AlfredPros: CodeLLaMa 7B Instruct Solidity",+ "provider": "alfredpros",+ "local_id": "codellama-7b-instruct-solidity",+ "context_length": 4096,+ "max_output_tokens": 4096,+ "input_price_per_mtok": 0.8,+ "output_price_per_mtok": 1.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "seed",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "xai/grok-3-mini-beta": {+ "id": "x-ai/grok-3-mini-beta",+ "name": "xAI: Grok 3 Mini Beta",+ "provider": "xai",+ "local_id": "grok-3-mini-beta",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.3,+ "output_price_per_mtok": 0.5,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "logprobs",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p"+ ]+ },+ "xai/grok-3-beta": {+ "id": "x-ai/grok-3-beta",+ "name": "xAI: Grok 3 Beta",+ "provider": "xai",+ "local_id": "grok-3-beta",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 3.0,+ "output_price_per_mtok": 15.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p"+ ]+ },+ "meta/llama-4-maverick": {+ "id": "meta-llama/llama-4-maverick",+ "name": "Meta: Llama 4 Maverick",+ "provider": "meta",+ "local_id": "llama-4-maverick",+ "context_length": 1048576,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 0.15,+ "output_price_per_mtok": 0.6,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "meta/llama-4-scout": {+ "id": "meta-llama/llama-4-scout",+ "name": "Meta: Llama 4 Scout",+ "provider": "meta",+ "local_id": "llama-4-scout",+ "context_length": 327680,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 0.08,+ "output_price_per_mtok": 0.3,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "qwen/qwen2.5-vl-32b-instruct": {+ "id": "qwen/qwen2.5-vl-32b-instruct",+ "name": "Qwen: Qwen2.5 VL 32B Instruct",+ "provider": "qwen",+ "local_id": "qwen2.5-vl-32b-instruct",+ "context_length": 128000,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.2,+ "output_price_per_mtok": 0.6,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "deepseek/deepseek-chat-v3-0324": {+ "id": "deepseek/deepseek-chat-v3-0324",+ "name": "DeepSeek: DeepSeek V3 0324",+ "provider": "deepseek",+ "local_id": "deepseek-chat-v3-0324",+ "context_length": 163840,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 0.19,+ "output_price_per_mtok": 0.87,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_logprobs",+ "top_p"+ ]+ },+ "openai/o1-pro": {+ "id": "openai/o1-pro",+ "name": "OpenAI: o1-pro",+ "provider": "openai",+ "local_id": "o1-pro",+ "context_length": 200000,+ "max_output_tokens": 100000,+ "input_price_per_mtok": 150.0,+ "output_price_per_mtok": 600.0,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 1.0,+ 1.0+ ],+ "is_reasoning_model": true,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "response_format",+ "seed",+ "structured_outputs"+ ]+ },+ "mistralai/mistral-small-3.1-24b-instruct:free": {+ "id": "mistralai/mistral-small-3.1-24b-instruct:free",+ "name": "Mistral: Mistral Small 3.1 24B (free)",+ "provider": "mistralai",+ "local_id": "mistral-small-3.1-24b-instruct:free",+ "context_length": 128000,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "mistralai/mistral-small-3.1-24b-instruct": {+ "id": "mistralai/mistral-small-3.1-24b-instruct",+ "name": "Mistral: Mistral Small 3.1 24B",+ "provider": "mistralai",+ "local_id": "mistral-small-3.1-24b-instruct",+ "context_length": 128000,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.35,+ "output_price_per_mtok": 0.56,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "repetition_penalty",+ "seed",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "allenai/olmo-2-0325-32b-instruct": {+ "id": "allenai/olmo-2-0325-32b-instruct",+ "name": "AllenAI: Olmo 2 32B Instruct",+ "provider": "allenai",+ "local_id": "olmo-2-0325-32b-instruct",+ "context_length": 128000,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.05,+ "output_price_per_mtok": 0.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": []+ },+ "gemini/gemma-3-4b-it:free": {+ "id": "google/gemma-3-4b-it:free",+ "name": "Google: Gemma 3 4B (free)",+ "provider": "gemini",+ "local_id": "gemma-3-4b-it:free",+ "context_length": 32768,+ "max_output_tokens": 8192,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "top_p"+ ]+ },+ "gemini/gemma-3-4b-it": {+ "id": "google/gemma-3-4b-it",+ "name": "Google: Gemma 3 4B",+ "provider": "gemini",+ "local_id": "gemma-3-4b-it",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.04,+ "output_price_per_mtok": 0.08,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "gemini/gemma-3-12b-it:free": {+ "id": "google/gemma-3-12b-it:free",+ "name": "Google: Gemma 3 12B (free)",+ "provider": "gemini",+ "local_id": "gemma-3-12b-it:free",+ "context_length": 32768,+ "max_output_tokens": 8192,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "seed",+ "stop",+ "temperature",+ "top_p"+ ]+ },+ "gemini/gemma-3-12b-it": {+ "id": "google/gemma-3-12b-it",+ "name": "Google: Gemma 3 12B",+ "provider": "gemini",+ "local_id": "gemma-3-12b-it",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.04,+ "output_price_per_mtok": 0.13,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "cohere/command-a": {+ "id": "cohere/command-a",+ "name": "Cohere: Command A",+ "provider": "cohere",+ "local_id": "command-a",+ "context_length": 256000,+ "max_output_tokens": 8192,+ "input_price_per_mtok": 2.5,+ "output_price_per_mtok": 10.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "openai/gpt-4o-mini-search-preview": {+ "id": "openai/gpt-4o-mini-search-preview",+ "name": "OpenAI: GPT-4o-mini Search Preview",+ "provider": "openai",+ "local_id": "gpt-4o-mini-search-preview",+ "context_length": 128000,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 0.15,+ "output_price_per_mtok": 0.6,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "response_format",+ "structured_outputs",+ "web_search_options"+ ]+ },+ "openai/gpt-4o-search-preview": {+ "id": "openai/gpt-4o-search-preview",+ "name": "OpenAI: GPT-4o Search Preview",+ "provider": "openai",+ "local_id": "gpt-4o-search-preview",+ "context_length": 128000,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 2.5,+ "output_price_per_mtok": 10.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "response_format",+ "structured_outputs",+ "web_search_options"+ ]+ },+ "gemini/gemma-3-27b-it:free": {+ "id": "google/gemma-3-27b-it:free",+ "name": "Google: Gemma 3 27B (free)",+ "provider": "gemini",+ "local_id": "gemma-3-27b-it:free",+ "context_length": 131072,+ "max_output_tokens": 8192,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "gemini/gemma-3-27b-it": {+ "id": "google/gemma-3-27b-it",+ "name": "Google: Gemma 3 27B",+ "provider": "gemini",+ "local_id": "gemma-3-27b-it",+ "context_length": 128000,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 0.04,+ "output_price_per_mtok": 0.15,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "thedrummer/skyfall-36b-v2": {+ "id": "thedrummer/skyfall-36b-v2",+ "name": "TheDrummer: Skyfall 36B V2",+ "provider": "thedrummer",+ "local_id": "skyfall-36b-v2",+ "context_length": 32768,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.55,+ "output_price_per_mtok": 0.8,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "presence_penalty",+ "repetition_penalty",+ "seed",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "perplexity/sonar-reasoning-pro": {+ "id": "perplexity/sonar-reasoning-pro",+ "name": "Perplexity: Sonar Reasoning Pro",+ "provider": "perplexity",+ "local_id": "sonar-reasoning-pro",+ "context_length": 128000,+ "max_output_tokens": 0,+ "input_price_per_mtok": 2.0,+ "output_price_per_mtok": 8.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "temperature",+ "top_k",+ "top_p",+ "web_search_options"+ ]+ },+ "perplexity/sonar-pro": {+ "id": "perplexity/sonar-pro",+ "name": "Perplexity: Sonar Pro",+ "provider": "perplexity",+ "local_id": "sonar-pro",+ "context_length": 200000,+ "max_output_tokens": 8000,+ "input_price_per_mtok": 3.0,+ "output_price_per_mtok": 15.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "temperature",+ "top_k",+ "top_p",+ "web_search_options"+ ]+ },+ "perplexity/sonar-deep-research": {+ "id": "perplexity/sonar-deep-research",+ "name": "Perplexity: Sonar Deep Research",+ "provider": "perplexity",+ "local_id": "sonar-deep-research",+ "context_length": 128000,+ "max_output_tokens": 0,+ "input_price_per_mtok": 2.0,+ "output_price_per_mtok": 8.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "temperature",+ "top_k",+ "top_p",+ "web_search_options"+ ]+ },+ "qwen/qwq-32b": {+ "id": "qwen/qwq-32b",+ "name": "Qwen: QwQ 32B",+ "provider": "qwen",+ "local_id": "qwq-32b",+ "context_length": 32768,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.15,+ "output_price_per_mtok": 0.4,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "response_format",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "gemini/gemini-2.0-flash-lite-001": {+ "id": "google/gemini-2.0-flash-lite-001",+ "name": "Google: Gemini 2.0 Flash Lite",+ "provider": "gemini",+ "local_id": "gemini-2.0-flash-lite-001",+ "context_length": 1048576,+ "max_output_tokens": 8192,+ "input_price_per_mtok": 0.075,+ "output_price_per_mtok": 0.3,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "anthropic/claude-3.7-sonnet": {+ "id": "anthropic/claude-3.7-sonnet",+ "name": "Anthropic: Claude 3.7 Sonnet",+ "provider": "anthropic",+ "local_id": "claude-3.7-sonnet",+ "context_length": 200000,+ "max_output_tokens": 64000,+ "input_price_per_mtok": 3.0,+ "output_price_per_mtok": 15.0,+ "supports_thinking": true,+ "thinking_type": "budget",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "anthropic/claude-3.7-sonnet:thinking": {+ "id": "anthropic/claude-3.7-sonnet:thinking",+ "name": "Anthropic: Claude 3.7 Sonnet (thinking)",+ "provider": "anthropic",+ "local_id": "claude-3.7-sonnet:thinking",+ "context_length": 200000,+ "max_output_tokens": 64000,+ "input_price_per_mtok": 3.0,+ "output_price_per_mtok": 15.0,+ "supports_thinking": true,+ "thinking_type": "budget",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "mistralai/mistral-saba": {+ "id": "mistralai/mistral-saba",+ "name": "Mistral: Saba",+ "provider": "mistralai",+ "local_id": "mistral-saba",+ "context_length": 32768,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.2,+ "output_price_per_mtok": 0.6,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "meta/llama-guard-3-8b": {+ "id": "meta-llama/llama-guard-3-8b",+ "name": "Llama Guard 3 8B",+ "provider": "meta",+ "local_id": "llama-guard-3-8b",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.02,+ "output_price_per_mtok": 0.06,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "repetition_penalty",+ "seed",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "openai/o3-mini-high": {+ "id": "openai/o3-mini-high",+ "name": "OpenAI: o3 Mini High",+ "provider": "openai",+ "local_id": "o3-mini-high",+ "context_length": 200000,+ "max_output_tokens": 100000,+ "input_price_per_mtok": 1.1,+ "output_price_per_mtok": 4.4,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 1.0,+ 1.0+ ],+ "is_reasoning_model": true,+ "supported_parameters": [+ "max_tokens",+ "response_format",+ "seed",+ "structured_outputs",+ "tool_choice",+ "tools"+ ]+ },+ "gemini/gemini-2.0-flash-001": {+ "id": "google/gemini-2.0-flash-001",+ "name": "Google: Gemini 2.0 Flash",+ "provider": "gemini",+ "local_id": "gemini-2.0-flash-001",+ "context_length": 1048576,+ "max_output_tokens": 8192,+ "input_price_per_mtok": 0.1,+ "output_price_per_mtok": 0.4,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "qwen/qwen-vl-plus": {+ "id": "qwen/qwen-vl-plus",+ "name": "Qwen: Qwen VL Plus",+ "provider": "qwen",+ "local_id": "qwen-vl-plus",+ "context_length": 131072,+ "max_output_tokens": 8192,+ "input_price_per_mtok": 0.21,+ "output_price_per_mtok": 0.63,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "temperature",+ "top_p"+ ]+ },+ "aion-labs/aion-1.0": {+ "id": "aion-labs/aion-1.0",+ "name": "AionLabs: Aion-1.0",+ "provider": "aion-labs",+ "local_id": "aion-1.0",+ "context_length": 131072,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 4.0,+ "output_price_per_mtok": 8.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "temperature",+ "top_p"+ ]+ },+ "aion-labs/aion-1.0-mini": {+ "id": "aion-labs/aion-1.0-mini",+ "name": "AionLabs: Aion-1.0-Mini",+ "provider": "aion-labs",+ "local_id": "aion-1.0-mini",+ "context_length": 131072,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.7,+ "output_price_per_mtok": 1.4,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_tokens",+ "reasoning",+ "temperature",+ "top_p"+ ]+ },+ "aion-labs/aion-rp-llama-3.1-8b": {+ "id": "aion-labs/aion-rp-llama-3.1-8b",+ "name": "AionLabs: Aion-RP 1.0 (8B)",+ "provider": "aion-labs",+ "local_id": "aion-rp-llama-3.1-8b",+ "context_length": 32768,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.8,+ "output_price_per_mtok": 1.6,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "temperature",+ "top_p"+ ]+ },+ "qwen/qwen-vl-max": {+ "id": "qwen/qwen-vl-max",+ "name": "Qwen: Qwen VL Max",+ "provider": "qwen",+ "local_id": "qwen-vl-max",+ "context_length": 131072,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.8,+ "output_price_per_mtok": 3.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "qwen/qwen-turbo": {+ "id": "qwen/qwen-turbo",+ "name": "Qwen: Qwen-Turbo",+ "provider": "qwen",+ "local_id": "qwen-turbo",+ "context_length": 131072,+ "max_output_tokens": 8192,+ "input_price_per_mtok": 0.05,+ "output_price_per_mtok": 0.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "qwen/qwen2.5-vl-72b-instruct": {+ "id": "qwen/qwen2.5-vl-72b-instruct",+ "name": "Qwen: Qwen2.5 VL 72B Instruct",+ "provider": "qwen",+ "local_id": "qwen2.5-vl-72b-instruct",+ "context_length": 32768,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.8,+ "output_price_per_mtok": 0.8,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "qwen/qwen-plus": {+ "id": "qwen/qwen-plus",+ "name": "Qwen: Qwen-Plus",+ "provider": "qwen",+ "local_id": "qwen-plus",+ "context_length": 1000000,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.4,+ "output_price_per_mtok": 1.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "qwen/qwen-max": {+ "id": "qwen/qwen-max",+ "name": "Qwen: Qwen-Max ",+ "provider": "qwen",+ "local_id": "qwen-max",+ "context_length": 32768,+ "max_output_tokens": 8192,+ "input_price_per_mtok": 1.6,+ "output_price_per_mtok": 6.4,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "openai/o3-mini": {+ "id": "openai/o3-mini",+ "name": "OpenAI: o3 Mini",+ "provider": "openai",+ "local_id": "o3-mini",+ "context_length": 200000,+ "max_output_tokens": 100000,+ "input_price_per_mtok": 1.1,+ "output_price_per_mtok": 4.4,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 1.0,+ 1.0+ ],+ "is_reasoning_model": true,+ "supported_parameters": [+ "max_tokens",+ "response_format",+ "seed",+ "structured_outputs",+ "tool_choice",+ "tools"+ ]+ },+ "mistralai/mistral-small-24b-instruct-2501": {+ "id": "mistralai/mistral-small-24b-instruct-2501",+ "name": "Mistral: Mistral Small 3",+ "provider": "mistralai",+ "local_id": "mistral-small-24b-instruct-2501",+ "context_length": 32768,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 0.05,+ "output_price_per_mtok": 0.08,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "deepseek/deepseek-r1-distill-qwen-32b": {+ "id": "deepseek/deepseek-r1-distill-qwen-32b",+ "name": "DeepSeek: R1 Distill Qwen 32B",+ "provider": "deepseek",+ "local_id": "deepseek-r1-distill-qwen-32b",+ "context_length": 32768,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.29,+ "output_price_per_mtok": 0.29,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "perplexity/sonar": {+ "id": "perplexity/sonar",+ "name": "Perplexity: Sonar",+ "provider": "perplexity",+ "local_id": "sonar",+ "context_length": 127072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 1.0,+ "output_price_per_mtok": 1.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "temperature",+ "top_k",+ "top_p",+ "web_search_options"+ ]+ },+ "deepseek/deepseek-r1-distill-llama-70b": {+ "id": "deepseek/deepseek-r1-distill-llama-70b",+ "name": "DeepSeek: R1 Distill Llama 70B",+ "provider": "deepseek",+ "local_id": "deepseek-r1-distill-llama-70b",+ "context_length": 131072,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 0.7,+ "output_price_per_mtok": 0.8,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "deepseek/deepseek-r1": {+ "id": "deepseek/deepseek-r1",+ "name": "DeepSeek: R1",+ "provider": "deepseek",+ "local_id": "deepseek-r1",+ "context_length": 64000,+ "max_output_tokens": 16000,+ "input_price_per_mtok": 0.7,+ "output_price_per_mtok": 2.5,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "seed",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "minimax/minimax-01": {+ "id": "minimax/minimax-01",+ "name": "MiniMax: MiniMax-01",+ "provider": "minimax",+ "local_id": "minimax-01",+ "context_length": 1000192,+ "max_output_tokens": 1000192,+ "input_price_per_mtok": 0.2,+ "output_price_per_mtok": 1.1,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "temperature",+ "top_p"+ ]+ },+ "microsoft/phi-4": {+ "id": "microsoft/phi-4",+ "name": "Microsoft: Phi 4",+ "provider": "microsoft",+ "local_id": "phi-4",+ "context_length": 16384,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 0.06,+ "output_price_per_mtok": 0.14,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "sao10k/l3.1-70b-hanami-x1": {+ "id": "sao10k/l3.1-70b-hanami-x1",+ "name": "Sao10K: Llama 3.1 70B Hanami x1",+ "provider": "sao10k",+ "local_id": "l3.1-70b-hanami-x1",+ "context_length": 16000,+ "max_output_tokens": 0,+ "input_price_per_mtok": 3.0,+ "output_price_per_mtok": 3.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "seed",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "deepseek/deepseek-chat": {+ "id": "deepseek/deepseek-chat",+ "name": "DeepSeek: DeepSeek V3",+ "provider": "deepseek",+ "local_id": "deepseek-chat",+ "context_length": 163840,+ "max_output_tokens": 163840,+ "input_price_per_mtok": 0.32,+ "output_price_per_mtok": 0.89,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "sao10k/l3.3-euryale-70b": {+ "id": "sao10k/l3.3-euryale-70b",+ "name": "Sao10K: Llama 3.3 Euryale 70B",+ "provider": "sao10k",+ "local_id": "l3.3-euryale-70b",+ "context_length": 131072,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 0.65,+ "output_price_per_mtok": 0.75,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "openai/o1": {+ "id": "openai/o1",+ "name": "OpenAI: o1",+ "provider": "openai",+ "local_id": "o1",+ "context_length": 200000,+ "max_output_tokens": 100000,+ "input_price_per_mtok": 15.0,+ "output_price_per_mtok": 60.0,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 1.0,+ 1.0+ ],+ "is_reasoning_model": true,+ "supported_parameters": [+ "max_tokens",+ "response_format",+ "seed",+ "structured_outputs",+ "tool_choice",+ "tools"+ ]+ },+ "cohere/command-r7b-12-2024": {+ "id": "cohere/command-r7b-12-2024",+ "name": "Cohere: Command R7B (12-2024)",+ "provider": "cohere",+ "local_id": "command-r7b-12-2024",+ "context_length": 128000,+ "max_output_tokens": 4000,+ "input_price_per_mtok": 0.0375,+ "output_price_per_mtok": 0.15,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "meta/llama-3.3-70b-instruct:free": {+ "id": "meta-llama/llama-3.3-70b-instruct:free",+ "name": "Meta: Llama 3.3 70B Instruct (free)",+ "provider": "meta",+ "local_id": "llama-3.3-70b-instruct:free",+ "context_length": 128000,+ "max_output_tokens": 128000,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "seed",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "meta/llama-3.3-70b-instruct": {+ "id": "meta-llama/llama-3.3-70b-instruct",+ "name": "Meta: Llama 3.3 70B Instruct",+ "provider": "meta",+ "local_id": "llama-3.3-70b-instruct",+ "context_length": 131072,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 0.1,+ "output_price_per_mtok": 0.32,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_logprobs",+ "top_p"+ ]+ },+ "amazon/nova-lite-v1": {+ "id": "amazon/nova-lite-v1",+ "name": "Amazon: Nova Lite 1.0",+ "provider": "amazon",+ "local_id": "nova-lite-v1",+ "context_length": 300000,+ "max_output_tokens": 5120,+ "input_price_per_mtok": 0.06,+ "output_price_per_mtok": 0.24,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "stop",+ "temperature",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "amazon/nova-micro-v1": {+ "id": "amazon/nova-micro-v1",+ "name": "Amazon: Nova Micro 1.0",+ "provider": "amazon",+ "local_id": "nova-micro-v1",+ "context_length": 128000,+ "max_output_tokens": 5120,+ "input_price_per_mtok": 0.035,+ "output_price_per_mtok": 0.14,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "stop",+ "temperature",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "amazon/nova-pro-v1": {+ "id": "amazon/nova-pro-v1",+ "name": "Amazon: Nova Pro 1.0",+ "provider": "amazon",+ "local_id": "nova-pro-v1",+ "context_length": 300000,+ "max_output_tokens": 5120,+ "input_price_per_mtok": 0.8,+ "output_price_per_mtok": 3.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "stop",+ "temperature",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "openai/gpt-4o-2024-11-20": {+ "id": "openai/gpt-4o-2024-11-20",+ "name": "OpenAI: GPT-4o (2024-11-20)",+ "provider": "openai",+ "local_id": "gpt-4o-2024-11-20",+ "context_length": 128000,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 2.5,+ "output_price_per_mtok": 10.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p",+ "web_search_options"+ ]+ },+ "mistralai/mistral-large-2411": {+ "id": "mistralai/mistral-large-2411",+ "name": "Mistral Large 2411",+ "provider": "mistralai",+ "local_id": "mistral-large-2411",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 2.0,+ "output_price_per_mtok": 6.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "mistralai/mistral-large-2407": {+ "id": "mistralai/mistral-large-2407",+ "name": "Mistral Large 2407",+ "provider": "mistralai",+ "local_id": "mistral-large-2407",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 2.0,+ "output_price_per_mtok": 6.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "mistralai/pixtral-large-2411": {+ "id": "mistralai/pixtral-large-2411",+ "name": "Mistral: Pixtral Large 2411",+ "provider": "mistralai",+ "local_id": "pixtral-large-2411",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 2.0,+ "output_price_per_mtok": 6.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "qwen/qwen-2.5-coder-32b-instruct": {+ "id": "qwen/qwen-2.5-coder-32b-instruct",+ "name": "Qwen2.5 Coder 32B Instruct",+ "provider": "qwen",+ "local_id": "qwen-2.5-coder-32b-instruct",+ "context_length": 32768,+ "max_output_tokens": 8192,+ "input_price_per_mtok": 0.2,+ "output_price_per_mtok": 0.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "seed",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "raifle/sorcererlm-8x22b": {+ "id": "raifle/sorcererlm-8x22b",+ "name": "SorcererLM 8x22B",+ "provider": "raifle",+ "local_id": "sorcererlm-8x22b",+ "context_length": 16000,+ "max_output_tokens": 0,+ "input_price_per_mtok": 4.5,+ "output_price_per_mtok": 4.5,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "seed",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "thedrummer/unslopnemo-12b": {+ "id": "thedrummer/unslopnemo-12b",+ "name": "TheDrummer: UnslopNemo 12B",+ "provider": "thedrummer",+ "local_id": "unslopnemo-12b",+ "context_length": 32768,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.4,+ "output_price_per_mtok": 0.4,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "anthropic/claude-3.5-haiku": {+ "id": "anthropic/claude-3.5-haiku",+ "name": "Anthropic: Claude 3.5 Haiku",+ "provider": "anthropic",+ "local_id": "claude-3.5-haiku",+ "context_length": 200000,+ "max_output_tokens": 8192,+ "input_price_per_mtok": 0.8,+ "output_price_per_mtok": 4.0,+ "supports_thinking": true,+ "thinking_type": "budget",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "anthracite-org/magnum-v4-72b": {+ "id": "anthracite-org/magnum-v4-72b",+ "name": "Magnum v4 72B",+ "provider": "anthracite-org",+ "local_id": "magnum-v4-72b",+ "context_length": 16384,+ "max_output_tokens": 2048,+ "input_price_per_mtok": 3.0,+ "output_price_per_mtok": 5.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "top_a",+ "top_k",+ "top_logprobs",+ "top_p"+ ]+ },+ "anthropic/claude-3.5-sonnet": {+ "id": "anthropic/claude-3.5-sonnet",+ "name": "Anthropic: Claude 3.5 Sonnet",+ "provider": "anthropic",+ "local_id": "claude-3.5-sonnet",+ "context_length": 200000,+ "max_output_tokens": 8192,+ "input_price_per_mtok": 6.0,+ "output_price_per_mtok": 30.0,+ "supports_thinking": true,+ "thinking_type": "budget",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "qwen/qwen-2.5-7b-instruct": {+ "id": "qwen/qwen-2.5-7b-instruct",+ "name": "Qwen: Qwen2.5 7B Instruct",+ "provider": "qwen",+ "local_id": "qwen-2.5-7b-instruct",+ "context_length": 32768,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.04,+ "output_price_per_mtok": 0.1,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "nvidia/llama-3.1-nemotron-70b-instruct": {+ "id": "nvidia/llama-3.1-nemotron-70b-instruct",+ "name": "NVIDIA: Llama 3.1 Nemotron 70B Instruct",+ "provider": "nvidia",+ "local_id": "llama-3.1-nemotron-70b-instruct",+ "context_length": 131072,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 1.2,+ "output_price_per_mtok": 1.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "inflection/inflection-3-pi": {+ "id": "inflection/inflection-3-pi",+ "name": "Inflection: Inflection 3 Pi",+ "provider": "inflection",+ "local_id": "inflection-3-pi",+ "context_length": 8000,+ "max_output_tokens": 1024,+ "input_price_per_mtok": 2.5,+ "output_price_per_mtok": 10.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "stop",+ "temperature",+ "top_p"+ ]+ },+ "inflection/inflection-3-productivity": {+ "id": "inflection/inflection-3-productivity",+ "name": "Inflection: Inflection 3 Productivity",+ "provider": "inflection",+ "local_id": "inflection-3-productivity",+ "context_length": 8000,+ "max_output_tokens": 1024,+ "input_price_per_mtok": 2.5,+ "output_price_per_mtok": 10.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "stop",+ "temperature",+ "top_p"+ ]+ },+ "thedrummer/rocinante-12b": {+ "id": "thedrummer/rocinante-12b",+ "name": "TheDrummer: Rocinante 12B",+ "provider": "thedrummer",+ "local_id": "rocinante-12b",+ "context_length": 32768,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.17,+ "output_price_per_mtok": 0.43,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "meta/llama-3.2-3b-instruct:free": {+ "id": "meta-llama/llama-3.2-3b-instruct:free",+ "name": "Meta: Llama 3.2 3B Instruct (free)",+ "provider": "meta",+ "local_id": "llama-3.2-3b-instruct:free",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "meta/llama-3.2-3b-instruct": {+ "id": "meta-llama/llama-3.2-3b-instruct",+ "name": "Meta: Llama 3.2 3B Instruct",+ "provider": "meta",+ "local_id": "llama-3.2-3b-instruct",+ "context_length": 131072,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 0.02,+ "output_price_per_mtok": 0.02,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "meta/llama-3.2-1b-instruct": {+ "id": "meta-llama/llama-3.2-1b-instruct",+ "name": "Meta: Llama 3.2 1B Instruct",+ "provider": "meta",+ "local_id": "llama-3.2-1b-instruct",+ "context_length": 60000,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.027,+ "output_price_per_mtok": 0.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "repetition_penalty",+ "seed",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "meta/llama-3.2-11b-vision-instruct": {+ "id": "meta-llama/llama-3.2-11b-vision-instruct",+ "name": "Meta: Llama 3.2 11B Vision Instruct",+ "provider": "meta",+ "local_id": "llama-3.2-11b-vision-instruct",+ "context_length": 131072,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 0.049,+ "output_price_per_mtok": 0.049,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "qwen/qwen-2.5-72b-instruct": {+ "id": "qwen/qwen-2.5-72b-instruct",+ "name": "Qwen2.5 72B Instruct",+ "provider": "qwen",+ "local_id": "qwen-2.5-72b-instruct",+ "context_length": 32768,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 0.12,+ "output_price_per_mtok": 0.39,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "neversleep/llama-3.1-lumimaid-8b": {+ "id": "neversleep/llama-3.1-lumimaid-8b",+ "name": "NeverSleep: Lumimaid v0.2 8B",+ "provider": "neversleep",+ "local_id": "llama-3.1-lumimaid-8b",+ "context_length": 32768,+ "max_output_tokens": 4096,+ "input_price_per_mtok": 0.09,+ "output_price_per_mtok": 0.6,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "stop",+ "structured_outputs",+ "temperature",+ "top_p"+ ]+ },+ "cohere/command-r-08-2024": {+ "id": "cohere/command-r-08-2024",+ "name": "Cohere: Command R (08-2024)",+ "provider": "cohere",+ "local_id": "command-r-08-2024",+ "context_length": 128000,+ "max_output_tokens": 4000,+ "input_price_per_mtok": 0.15,+ "output_price_per_mtok": 0.6,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "cohere/command-r-plus-08-2024": {+ "id": "cohere/command-r-plus-08-2024",+ "name": "Cohere: Command R+ (08-2024)",+ "provider": "cohere",+ "local_id": "command-r-plus-08-2024",+ "context_length": 128000,+ "max_output_tokens": 4000,+ "input_price_per_mtok": 2.5,+ "output_price_per_mtok": 10.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "sao10k/l3.1-euryale-70b": {+ "id": "sao10k/l3.1-euryale-70b",+ "name": "Sao10K: Llama 3.1 Euryale 70B v2.2",+ "provider": "sao10k",+ "local_id": "l3.1-euryale-70b",+ "context_length": 32768,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.65,+ "output_price_per_mtok": 0.75,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "qwen/qwen-2.5-vl-7b-instruct": {+ "id": "qwen/qwen-2.5-vl-7b-instruct",+ "name": "Qwen: Qwen2.5-VL 7B Instruct",+ "provider": "qwen",+ "local_id": "qwen-2.5-vl-7b-instruct",+ "context_length": 32768,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.2,+ "output_price_per_mtok": 0.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "seed",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "nousresearch/hermes-3-llama-3.1-70b": {+ "id": "nousresearch/hermes-3-llama-3.1-70b",+ "name": "Nous: Hermes 3 70B Instruct",+ "provider": "nousresearch",+ "local_id": "hermes-3-llama-3.1-70b",+ "context_length": 65536,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 0.3,+ "output_price_per_mtok": 0.3,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "nousresearch/hermes-3-llama-3.1-405b:free": {+ "id": "nousresearch/hermes-3-llama-3.1-405b:free",+ "name": "Nous: Hermes 3 405B Instruct (free)",+ "provider": "nousresearch",+ "local_id": "hermes-3-llama-3.1-405b:free",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.0,+ "output_price_per_mtok": 0.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "nousresearch/hermes-3-llama-3.1-405b": {+ "id": "nousresearch/hermes-3-llama-3.1-405b",+ "name": "Nous: Hermes 3 405B Instruct",+ "provider": "nousresearch",+ "local_id": "hermes-3-llama-3.1-405b",+ "context_length": 131072,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 1.0,+ "output_price_per_mtok": 1.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "sao10k/l3-lunaris-8b": {+ "id": "sao10k/l3-lunaris-8b",+ "name": "Sao10K: Llama 3 8B Lunaris",+ "provider": "sao10k",+ "local_id": "l3-lunaris-8b",+ "context_length": 8192,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.04,+ "output_price_per_mtok": 0.05,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "openai/gpt-4o-2024-08-06": {+ "id": "openai/gpt-4o-2024-08-06",+ "name": "OpenAI: GPT-4o (2024-08-06)",+ "provider": "openai",+ "local_id": "gpt-4o-2024-08-06",+ "context_length": 128000,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 2.5,+ "output_price_per_mtok": 10.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p",+ "web_search_options"+ ]+ },+ "meta/llama-3.1-405b": {+ "id": "meta-llama/llama-3.1-405b",+ "name": "Meta: Llama 3.1 405B (base)",+ "provider": "meta",+ "local_id": "llama-3.1-405b",+ "context_length": 32768,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 4.0,+ "output_price_per_mtok": 4.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "seed",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "meta/llama-3.1-8b-instruct": {+ "id": "meta-llama/llama-3.1-8b-instruct",+ "name": "Meta: Llama 3.1 8B Instruct",+ "provider": "meta",+ "local_id": "llama-3.1-8b-instruct",+ "context_length": 16384,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 0.02,+ "output_price_per_mtok": 0.05,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_logprobs",+ "top_p"+ ]+ },+ "meta/llama-3.1-405b-instruct": {+ "id": "meta-llama/llama-3.1-405b-instruct",+ "name": "Meta: Llama 3.1 405B Instruct",+ "provider": "meta",+ "local_id": "llama-3.1-405b-instruct",+ "context_length": 131000,+ "max_output_tokens": 0,+ "input_price_per_mtok": 4.0,+ "output_price_per_mtok": 4.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "meta/llama-3.1-70b-instruct": {+ "id": "meta-llama/llama-3.1-70b-instruct",+ "name": "Meta: Llama 3.1 70B Instruct",+ "provider": "meta",+ "local_id": "llama-3.1-70b-instruct",+ "context_length": 131072,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.4,+ "output_price_per_mtok": 0.4,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "mistralai/mistral-nemo": {+ "id": "mistralai/mistral-nemo",+ "name": "Mistral: Mistral Nemo",+ "provider": "mistralai",+ "local_id": "mistral-nemo",+ "context_length": 131072,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 0.02,+ "output_price_per_mtok": 0.04,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "openai/gpt-4o-mini-2024-07-18": {+ "id": "openai/gpt-4o-mini-2024-07-18",+ "name": "OpenAI: GPT-4o-mini (2024-07-18)",+ "provider": "openai",+ "local_id": "gpt-4o-mini-2024-07-18",+ "context_length": 128000,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 0.15,+ "output_price_per_mtok": 0.6,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p",+ "web_search_options"+ ]+ },+ "openai/gpt-4o-mini": {+ "id": "openai/gpt-4o-mini",+ "name": "OpenAI: GPT-4o-mini",+ "provider": "openai",+ "local_id": "gpt-4o-mini",+ "context_length": 128000,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 0.15,+ "output_price_per_mtok": 0.6,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p",+ "web_search_options"+ ]+ },+ "gemini/gemma-2-27b-it": {+ "id": "google/gemma-2-27b-it",+ "name": "Google: Gemma 2 27B",+ "provider": "gemini",+ "local_id": "gemma-2-27b-it",+ "context_length": 8192,+ "max_output_tokens": 2048,+ "input_price_per_mtok": 0.65,+ "output_price_per_mtok": 0.65,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "stop",+ "structured_outputs",+ "temperature",+ "top_p"+ ]+ },+ "gemini/gemma-2-9b-it": {+ "id": "google/gemma-2-9b-it",+ "name": "Google: Gemma 2 9B",+ "provider": "gemini",+ "local_id": "gemma-2-9b-it",+ "context_length": 8192,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.03,+ "output_price_per_mtok": 0.09,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "repetition_penalty",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "sao10k/l3-euryale-70b": {+ "id": "sao10k/l3-euryale-70b",+ "name": "Sao10k: Llama 3 Euryale 70B v2.1",+ "provider": "sao10k",+ "local_id": "l3-euryale-70b",+ "context_length": 8192,+ "max_output_tokens": 8192,+ "input_price_per_mtok": 1.48,+ "output_price_per_mtok": 1.48,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "repetition_penalty",+ "seed",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "nousresearch/hermes-2-pro-llama-3-8b": {+ "id": "nousresearch/hermes-2-pro-llama-3-8b",+ "name": "NousResearch: Hermes 2 Pro - Llama-3 8B",+ "provider": "nousresearch",+ "local_id": "hermes-2-pro-llama-3-8b",+ "context_length": 8192,+ "max_output_tokens": 8192,+ "input_price_per_mtok": 0.14,+ "output_price_per_mtok": 0.14,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "mistralai/mistral-7b-instruct": {+ "id": "mistralai/mistral-7b-instruct",+ "name": "Mistral: Mistral 7B Instruct",+ "provider": "mistralai",+ "local_id": "mistral-7b-instruct",+ "context_length": 32768,+ "max_output_tokens": 4096,+ "input_price_per_mtok": 0.2,+ "output_price_per_mtok": 0.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "mistralai/mistral-7b-instruct-v0.3": {+ "id": "mistralai/mistral-7b-instruct-v0.3",+ "name": "Mistral: Mistral 7B Instruct v0.3",+ "provider": "mistralai",+ "local_id": "mistral-7b-instruct-v0.3",+ "context_length": 32768,+ "max_output_tokens": 4096,+ "input_price_per_mtok": 0.2,+ "output_price_per_mtok": 0.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "meta/llama-guard-2-8b": {+ "id": "meta-llama/llama-guard-2-8b",+ "name": "Meta: LlamaGuard 2 8B",+ "provider": "meta",+ "local_id": "llama-guard-2-8b",+ "context_length": 8192,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.2,+ "output_price_per_mtok": 0.2,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "openai/gpt-4o-2024-05-13": {+ "id": "openai/gpt-4o-2024-05-13",+ "name": "OpenAI: GPT-4o (2024-05-13)",+ "provider": "openai",+ "local_id": "gpt-4o-2024-05-13",+ "context_length": 128000,+ "max_output_tokens": 4096,+ "input_price_per_mtok": 5.0,+ "output_price_per_mtok": 15.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p",+ "web_search_options"+ ]+ },+ "openai/gpt-4o": {+ "id": "openai/gpt-4o",+ "name": "OpenAI: GPT-4o",+ "provider": "openai",+ "local_id": "gpt-4o",+ "context_length": 128000,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 2.5,+ "output_price_per_mtok": 10.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p",+ "web_search_options"+ ]+ },+ "openai/gpt-4o:extended": {+ "id": "openai/gpt-4o:extended",+ "name": "OpenAI: GPT-4o (extended)",+ "provider": "openai",+ "local_id": "gpt-4o:extended",+ "context_length": 128000,+ "max_output_tokens": 64000,+ "input_price_per_mtok": 6.0,+ "output_price_per_mtok": 18.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p",+ "web_search_options"+ ]+ },+ "meta/llama-3-70b-instruct": {+ "id": "meta-llama/llama-3-70b-instruct",+ "name": "Meta: Llama 3 70B Instruct",+ "provider": "meta",+ "local_id": "llama-3-70b-instruct",+ "context_length": 8192,+ "max_output_tokens": 8000,+ "input_price_per_mtok": 0.51,+ "output_price_per_mtok": 0.74,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "meta/llama-3-8b-instruct": {+ "id": "meta-llama/llama-3-8b-instruct",+ "name": "Meta: Llama 3 8B Instruct",+ "provider": "meta",+ "local_id": "llama-3-8b-instruct",+ "context_length": 8192,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 0.03,+ "output_price_per_mtok": 0.04,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "mistralai/mixtral-8x22b-instruct": {+ "id": "mistralai/mixtral-8x22b-instruct",+ "name": "Mistral: Mixtral 8x22B Instruct",+ "provider": "mistralai",+ "local_id": "mixtral-8x22b-instruct",+ "context_length": 65536,+ "max_output_tokens": 0,+ "input_price_per_mtok": 2.0,+ "output_price_per_mtok": 6.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "microsoft/wizardlm-2-8x22b": {+ "id": "microsoft/wizardlm-2-8x22b",+ "name": "WizardLM-2 8x22B",+ "provider": "microsoft",+ "local_id": "wizardlm-2-8x22b",+ "context_length": 65535,+ "max_output_tokens": 8000,+ "input_price_per_mtok": 0.62,+ "output_price_per_mtok": 0.62,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "repetition_penalty",+ "seed",+ "stop",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "openai/gpt-4-turbo": {+ "id": "openai/gpt-4-turbo",+ "name": "OpenAI: GPT-4 Turbo",+ "provider": "openai",+ "local_id": "gpt-4-turbo",+ "context_length": 128000,+ "max_output_tokens": 4096,+ "input_price_per_mtok": 10.0,+ "output_price_per_mtok": 30.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p"+ ]+ },+ "anthropic/claude-3-haiku": {+ "id": "anthropic/claude-3-haiku",+ "name": "Anthropic: Claude 3 Haiku",+ "provider": "anthropic",+ "local_id": "claude-3-haiku",+ "context_length": 200000,+ "max_output_tokens": 4096,+ "input_price_per_mtok": 0.25,+ "output_price_per_mtok": 1.25,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 1.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "max_tokens",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "mistralai/mistral-large": {+ "id": "mistralai/mistral-large",+ "name": "Mistral Large",+ "provider": "mistralai",+ "local_id": "mistral-large",+ "context_length": 128000,+ "max_output_tokens": 0,+ "input_price_per_mtok": 2.0,+ "output_price_per_mtok": 6.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },+ "openai/gpt-3.5-turbo-0613": {+ "id": "openai/gpt-3.5-turbo-0613",+ "name": "OpenAI: GPT-3.5 Turbo (older v0613)",+ "provider": "openai",+ "local_id": "gpt-3.5-turbo-0613",+ "context_length": 4095,+ "max_output_tokens": 4096,+ "input_price_per_mtok": 1.0,+ "output_price_per_mtok": 2.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p"+ ]+ },+ "openai/gpt-4-turbo-preview": {+ "id": "openai/gpt-4-turbo-preview",+ "name": "OpenAI: GPT-4 Turbo Preview",+ "provider": "openai",+ "local_id": "gpt-4-turbo-preview",+ "context_length": 128000,+ "max_output_tokens": 4096,+ "input_price_per_mtok": 10.0,+ "output_price_per_mtok": 30.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p"+ ]+ },+ "mistralai/mixtral-8x7b-instruct": {+ "id": "mistralai/mixtral-8x7b-instruct",+ "name": "Mistral: Mixtral 8x7B Instruct",+ "provider": "mistralai",+ "local_id": "mixtral-8x7b-instruct",+ "context_length": 32768,+ "max_output_tokens": 16384,+ "input_price_per_mtok": 0.54,+ "output_price_per_mtok": 0.54,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_p"+ ]+ },+ "neversleep/noromaid-20b": {+ "id": "neversleep/noromaid-20b",+ "name": "Noromaid 20B",+ "provider": "neversleep",+ "local_id": "noromaid-20b",+ "context_length": 4096,+ "max_output_tokens": 2048,+ "input_price_per_mtok": 1.0,+ "output_price_per_mtok": 1.75,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "stop",+ "structured_outputs",+ "temperature",+ "top_p"+ ]+ },+ "alpindale/goliath-120b": {+ "id": "alpindale/goliath-120b",+ "name": "Goliath 120B",+ "provider": "alpindale",+ "local_id": "goliath-120b",+ "context_length": 6144,+ "max_output_tokens": 1024,+ "input_price_per_mtok": 3.75,+ "output_price_per_mtok": 7.5,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "top_a",+ "top_k",+ "top_logprobs",+ "top_p"+ ]+ },+ "openrouter/auto": {+ "id": "openrouter/auto",+ "name": "Auto Router",+ "provider": "openrouter",+ "local_id": "auto",+ "context_length": 2000000,+ "max_output_tokens": 0,+ "input_price_per_mtok": -1000000.0,+ "output_price_per_mtok": -1000000.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "include_reasoning",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "reasoning",+ "reasoning_effort",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_logprobs",+ "top_p",+ "web_search_options"+ ]+ },+ "openai/gpt-4-1106-preview": {+ "id": "openai/gpt-4-1106-preview",+ "name": "OpenAI: GPT-4 Turbo (older v1106)",+ "provider": "openai",+ "local_id": "gpt-4-1106-preview",+ "context_length": 128000,+ "max_output_tokens": 4096,+ "input_price_per_mtok": 10.0,+ "output_price_per_mtok": 30.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p"+ ]+ },+ "openai/gpt-3.5-turbo-instruct": {+ "id": "openai/gpt-3.5-turbo-instruct",+ "name": "OpenAI: GPT-3.5 Turbo Instruct",+ "provider": "openai",+ "local_id": "gpt-3.5-turbo-instruct",+ "context_length": 4095,+ "max_output_tokens": 4096,+ "input_price_per_mtok": 1.5,+ "output_price_per_mtok": 2.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_logprobs",+ "top_p"+ ]+ },+ "mistralai/mistral-7b-instruct-v0.1": {+ "id": "mistralai/mistral-7b-instruct-v0.1",+ "name": "Mistral: Mistral 7B Instruct v0.1",+ "provider": "mistralai",+ "local_id": "mistral-7b-instruct-v0.1",+ "context_length": 2824,+ "max_output_tokens": 0,+ "input_price_per_mtok": 0.11,+ "output_price_per_mtok": 0.19,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "presence_penalty",+ "repetition_penalty",+ "seed",+ "temperature",+ "top_k",+ "top_p"+ ]+ },+ "openai/gpt-3.5-turbo-16k": {+ "id": "openai/gpt-3.5-turbo-16k",+ "name": "OpenAI: GPT-3.5 Turbo 16k",+ "provider": "openai",+ "local_id": "gpt-3.5-turbo-16k",+ "context_length": 16385,+ "max_output_tokens": 4096,+ "input_price_per_mtok": 3.0,+ "output_price_per_mtok": 4.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p"+ ]+ },+ "mancer/weaver": {+ "id": "mancer/weaver",+ "name": "Mancer: Weaver (alpha)",+ "provider": "mancer",+ "local_id": "weaver",+ "context_length": 8000,+ "max_output_tokens": 2000,+ "input_price_per_mtok": 0.75,+ "output_price_per_mtok": 1.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "temperature",+ "top_a",+ "top_k",+ "top_logprobs",+ "top_p"+ ]+ },+ "undi95/remm-slerp-l2-13b": {+ "id": "undi95/remm-slerp-l2-13b",+ "name": "ReMM SLERP 13B",+ "provider": "undi95",+ "local_id": "remm-slerp-l2-13b",+ "context_length": 6144,+ "max_output_tokens": 4096,+ "input_price_per_mtok": 0.45,+ "output_price_per_mtok": 0.65,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_a",+ "top_k",+ "top_logprobs",+ "top_p"+ ]+ },+ "gryphe/mythomax-l2-13b": {+ "id": "gryphe/mythomax-l2-13b",+ "name": "MythoMax 13B",+ "provider": "gryphe",+ "local_id": "mythomax-l2-13b",+ "context_length": 4096,+ "max_output_tokens": 4096,+ "input_price_per_mtok": 0.06,+ "output_price_per_mtok": 0.06,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "min_p",+ "presence_penalty",+ "repetition_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "top_a",+ "top_k",+ "top_logprobs",+ "top_p"+ ]+ },+ "openai/gpt-4-0314": {+ "id": "openai/gpt-4-0314",+ "name": "OpenAI: GPT-4 (older v0314)",+ "provider": "openai",+ "local_id": "gpt-4-0314",+ "context_length": 8191,+ "max_output_tokens": 4096,+ "input_price_per_mtok": 30.0,+ "output_price_per_mtok": 60.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p"+ ]+ },+ "openai/gpt-4": {+ "id": "openai/gpt-4",+ "name": "OpenAI: GPT-4",+ "provider": "openai",+ "local_id": "gpt-4",+ "context_length": 8191,+ "max_output_tokens": 4096,+ "input_price_per_mtok": 30.0,+ "output_price_per_mtok": 60.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p"+ ]+ },+ "openai/gpt-3.5-turbo": {+ "id": "openai/gpt-3.5-turbo",+ "name": "OpenAI: GPT-3.5 Turbo",+ "provider": "openai",+ "local_id": "gpt-3.5-turbo",+ "context_length": 16385,+ "max_output_tokens": 4096,+ "input_price_per_mtok": 0.5,+ "output_price_per_mtok": 1.5,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "logit_bias",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_logprobs",+ "top_p"+ ]+ }+ }+}\ No newline at end of filediff --git a/server/constants.py b/server/constants.pyindex 2c8ab260..802d573b 100644--- a/server/constants.py+++ b/server/constants.py@@ -283,6 +283,7 @@ WORKFLOW_TRIGGER_TYPES: FrozenSet[str] = frozenset(['taskTrigger','twitterReceive','gmailReceive',+ 'telegramReceive',])# =============================================================================diff --git a/server/main.py b/server/main.pyindex 3d18612d..3c7c7f88 100644--- a/server/main.py+++ b/server/main.py@@ -134,6 +134,22 @@ async def lifespan(app: FastAPI):compaction_svc.set_ai_service(container.ai_service())logger.info("Compaction service initialized")+ # Initialize model registry service+ from services.model_registry import get_model_registry+ model_registry = get_model_registry()+ model_registry.startup()+ logger.info("Model registry initialized")++ # Background refresh if cache is stale+ if model_registry.is_stale():+ async def _refresh_registry():+ try:+ count = await model_registry.refresh()+ logger.info(f"Model registry refreshed: {count} models")+ except Exception as e:+ logger.warning(f"Model registry refresh failed (offline?): {e}")+ asyncio.create_task(_refresh_registry())+# Initialize agent team servicefrom services.agent_team import init_agent_team_servicefrom services.status_broadcaster import get_status_broadcasterdiff --git a/server/routers/websocket.py b/server/routers/websocket.pyindex b18ffc00..1fa1b883 100644--- a/server/routers/websocket.py+++ b/server/routers/websocket.py@@ -1803,6 +1803,10 @@ async def handle_telegram_connect(data: Dict[str, Any], websocket: WebSocket) ->models=[],session_id="default",)+ # Restore owner_chat_id from credentials if previously captured+ saved_owner = await auth_service.get_api_key("telegram_owner_chat_id")+ if saved_owner:+ await service.set_owner(int(saved_owner))except Exception as e:logger.warning(f"[Telegram] Failed to store token in credentials: {e}")@@ -1926,6 +1930,16 @@ async def handle_telegram_reconnect(data: Dict[str, Any], websocket: WebSocket)service = get_telegram_service()result = await service.connect(token)++ # Restore owner_chat_id from credentials if previously captured+ if result.get("success"):+ try:+ saved_owner = await auth_service.get_api_key("telegram_owner_chat_id")+ if saved_owner:+ await service.set_owner(int(saved_owner))+ except Exception as e:+ logger.warning(f"[Telegram] Failed to restore owner_chat_id: {e}")+return resultexcept Exception as e:logger.error(f"[Telegram] Reconnect failed: {e}")@@ -2637,12 +2651,19 @@ async def handle_get_api_usage_summary(data: Dict[str, Any], websocket: WebSocke@ws_handler("session_id")async def handle_get_compaction_stats(data: Dict[str, Any], websocket: WebSocket) -> Dict[str, Any]:- """Get compaction statistics for a session."""+ """Get compaction statistics for a session.++ Optional model/provider params enable model-aware threshold (50% of context window).+ """from services.compaction import get_compaction_servicesvc = get_compaction_service()if not svc:return {"success": False, "error": "Compaction service not initialized"}- return await svc.stats(data["session_id"])+ return await svc.stats(+ data["session_id"],+ model=data.get("model", ""),+ provider=data.get("provider", ""),+ )@ws_handler("session_id")@@ -2799,6 +2820,35 @@ async def handle_get_team_messages(data: Dict[str, Any], websocket: WebSocket) -return {"messages": messages}+# ============================================================================+# Model Registry Handlers+# ============================================================================++@ws_handler()+async def handle_get_model_constraints(data: Dict[str, Any], websocket: WebSocket) -> Dict[str, Any]:+ """Get model constraints (max_output_tokens, temperature range, thinking support, etc.)."""+ from services.model_registry import get_model_registry+ registry = get_model_registry()+ model = data.get("model", "")+ provider = data.get("provider", "")+ if not model or not provider:+ return {"success": False, "error": "model and provider are required"}+ constraints = registry.get_model_constraints(model, provider)+ return {"success": True, **constraints}+++@ws_handler()+async def handle_refresh_model_registry(data: Dict[str, Any], websocket: WebSocket) -> Dict[str, Any]:+ """Force refresh model registry from OpenRouter."""+ from services.model_registry import get_model_registry+ registry = get_model_registry()+ try:+ count = await registry.refresh()+ return {"success": True, "model_count": count}+ except Exception as e:+ return {"success": False, "error": str(e)}++# ============================================================================# Message Router# ============================================================================@@ -2967,6 +3017,10 @@ MESSAGE_HANDLERS: Dict[str, MessageHandler] = {"save_pricing_config": handle_save_pricing_config,"get_api_usage_summary": handle_get_api_usage_summary,+ # Model Registry+ "get_model_constraints": handle_get_model_constraints,+ "refresh_model_registry": handle_refresh_model_registry,+# Agent Teams"create_team": handle_create_team,"get_team": handle_get_team,diff --git a/server/services/ai.py b/server/services/ai.pyindex 3f0ed577..e3758a60 100644--- a/server/services/ai.py+++ b/server/services/ai.py@@ -215,6 +215,55 @@ def _get_default_model(provider: str, fallback: str) -> str:return providers.get(provider, {}).get("default_model", fallback)+def _resolve_max_tokens(flattened: dict, model: str, provider: str) -> int:+ """Resolve max_tokens: user param -> model registry -> llm_defaults -> 4096.++ Clamps to model's actual maximum if user value exceeds it.+ """+ from services.model_registry import get_model_registry+ registry = get_model_registry()+ model_max = registry.get_max_output_tokens(model, provider)++ user_val = flattened.get('max_tokens') or flattened.get('maxTokens')+ if user_val:+ user_int = int(user_val)+ if user_int > model_max:+ logger.info(f"[AI] Clamping max_tokens from {user_int} to {model_max} for {provider}/{model}")+ return model_max+ return user_int+ return model_max+++def _resolve_temperature(flattened: dict, model: str, provider: str, thinking_enabled: bool) -> float:+ """Resolve temperature with model-specific constraints.++ Handles:+ - O-series models: always temp=1+ - Claude with thinking: always temp=1+ - Range clamping per provider+ """+ from services.model_registry import get_model_registry+ registry = get_model_registry()++ user_temp = float(flattened.get('temperature', 0.7))++ # O-series reasoning models always require temperature=1+ if registry.is_reasoning_model(model, provider):+ if user_temp != 1:+ logger.info(f"[AI] Reasoning model '{model}': forcing temperature to 1 (was {user_temp})")+ return 1.0++ # Claude thinking mode requires temperature=1+ if thinking_enabled and provider == 'anthropic':+ if user_temp != 1:+ logger.info(f"[AI] Claude thinking mode: forcing temperature to 1 (was {user_temp})")+ return 1.0++ # Clamp to valid range for provider/model+ temp_range = registry.get_temperature_range(model, provider)+ return max(temp_range[0], min(temp_range[1], user_temp))++# Provider configurations - defaults from config/llm_defaults.jsonPROVIDER_CONFIGS: Dict[str, ProviderConfig] = {'openai': ProviderConfig(@@ -263,7 +312,7 @@ PROVIDER_CONFIGS: Dict[str, ProviderConfig] = {api_key_param='api_key',max_tokens_param='max_tokens',detection_patterns=('groq',),- default_model=_get_default_model('groq', 'llama-4-maverick'),+ default_model=_get_default_model('groq', 'llama-4-scout'),models_endpoint='https://api.groq.com/openai/v1/models',models_header_fn=_groq_headers),@@ -880,17 +929,6 @@ class AIService:logger.warning(f"[LangGraph] Empty response with no error details. Content type: {type(content)}, value: {repr(content)}")raise ValueError("AI generated empty response. Try rephrasing your prompt or using a different model.")- def _is_reasoning_model(self, model: str) -> bool:- """Check if model supports reasoning (OpenAI o-series).-- O-series models: o1, o1-mini, o1-preview, o3, o3-mini, o4-mini, etc.- NOT gpt-4o (which contains 'o' but is not a reasoning model).- """- model_lower = model.lower()- # Check for o-series pattern: starts with 'o' followed by digit- # e.g., o1, o1-mini, o3, o3-mini, o4-mini- return bool(re.match(r'^o[134](-|$)', model_lower))-async def _track_token_usage(self,session_id: str,@@ -1075,77 +1113,47 @@ class AIService:'X-Title': 'MachinaOS'}- # OpenAI o-series reasoning models ALWAYS require temperature=1- # This applies regardless of whether thinking mode is enabled- if provider == 'openai' and self._is_reasoning_model(model):+ # Apply thinking/reasoning configuration using model registry+ # The registry determines thinking_type based on model metadata+ from services.model_registry import get_model_registry+ registry = get_model_registry()++ # Force temperature=1 for reasoning models (regardless of thinking mode)+ if registry.is_reasoning_model(model, provider):if kwargs.get('temperature', 1) != 1:- logger.info(f"[AI] OpenAI o-series model '{model}': forcing temperature to 1 (was {kwargs.get('temperature')})")+ logger.info(f"[AI] Reasoning model '{model}': forcing temperature to 1 (was {kwargs.get('temperature')})")kwargs['temperature'] = 1- # Apply thinking/reasoning configuration per provider (per LangChain docs Jan 2026)if thinking and thinking.enabled:- if provider == 'anthropic':- # Claude extended thinking: thinking={"type": "enabled", "budget_tokens": N}- # Requires temperature=1, budget min 1024 tokens- # IMPORTANT: max_tokens must be greater than budget_tokens+ thinking_type = registry.get_thinking_type(model, provider)++ if thinking_type == 'budget':+ # Claude, Gemini, Cerebras: budget_tokens approachbudget = max(1024, thinking.budget)- # Ensure max_tokens > budget_tokens (add buffer for response)- if max_tokens <= budget:- # Set max_tokens to budget + reasonable response space (at least 1024 more)- kwargs[config.max_tokens_param] = budget + max(1024, max_tokens)- logger.info(f"[AI] Claude thinking: adjusted max_tokens from {max_tokens} to {kwargs[config.max_tokens_param]} (budget={budget})")- kwargs['thinking'] = {"type": "enabled", "budget_tokens": budget}- kwargs['temperature'] = 1 # Required for Claude thinking mode- elif provider == 'openai' and self._is_reasoning_model(model):- # OpenAI o-series: Use reasoning_effort parameter- # Note: reasoning.summary requires organization verification on OpenAI- # So we just use reasoning_effort for now- kwargs['reasoning_effort'] = thinking.effort- # O-series models only support temperature=1- kwargs['temperature'] = 1- logger.info(f"[AI] OpenAI o-series: reasoning_effort={thinking.effort}, temperature=1")- elif provider == 'gemini':- # Gemini thinking support varies by model:- # - Gemini 2.5 Flash/Pro: Use thinking_budget (int tokens, 0=off, -1=dynamic)- # - Gemini 2.0 Flash Thinking: Built-in thinking, use thinking_budget- # - Gemini 3 models: Limited/experimental thinking support- model_lower = model.lower()-- # Gemini 2.5 models support thinking_budget- if 'gemini-2.5' in model_lower or '2.5' in model_lower:+ if provider == 'anthropic':+ # Claude: max_tokens must be > budget_tokens+ if max_tokens <= budget:+ kwargs[config.max_tokens_param] = budget + max(1024, max_tokens)+ logger.info(f"[AI] Claude thinking: adjusted max_tokens from {max_tokens} to {kwargs[config.max_tokens_param]} (budget={budget})")+ kwargs['thinking'] = {"type": "enabled", "budget_tokens": budget}+ kwargs['temperature'] = 1+ elif provider == 'gemini':kwargs['thinking_budget'] = thinking.budgetkwargs['include_thoughts'] = True- logger.info(f"[AI] Gemini 2.5 thinking: budget={thinking.budget}")- # Gemini 2.0 Flash Thinking model- elif 'gemini-2.0-flash-thinking' in model_lower or 'flash-thinking' in model_lower:+ logger.info(f"[AI] Gemini thinking: budget={thinking.budget}")+ elif provider == 'cerebras':kwargs['thinking_budget'] = thinking.budget- kwargs['include_thoughts'] = True- logger.info(f"[AI] Gemini Flash Thinking: budget={thinking.budget}")- # Gemini 3 preview models - thinking support is experimental/limited- # Only 'low' and 'high' levels may be supported, not 'medium'- elif 'gemini-3' in model_lower:- # For Gemini 3, try thinking_budget instead of thinking_level- # as level support is inconsistent across preview models- kwargs['thinking_budget'] = thinking.budget- kwargs['include_thoughts'] = True- logger.info(f"[AI] Gemini 3 thinking: using budget={thinking.budget} (level support varies)")- # Other Gemini 2.x models - try thinking_budget- elif 'gemini-2' in model_lower:- kwargs['thinking_budget'] = thinking.budget- kwargs['include_thoughts'] = True- logger.info(f"[AI] Gemini 2.x thinking: budget={thinking.budget}")- else:- # For other/older Gemini models, thinking may not be supported- logger.warning(f"[AI] Gemini model '{model}' may not support thinking mode")- elif provider == 'groq':- # Groq: reasoning_format ('parsed' or 'hidden')- # 'parsed' includes reasoning in additional_kwargs, 'hidden' suppresses it+ elif thinking_type == 'effort':+ # OpenAI o-series: reasoning_effort parameter+ kwargs['reasoning_effort'] = thinking.effort+ kwargs['temperature'] = 1+ logger.info(f"[AI] OpenAI reasoning: effort={thinking.effort}, temperature=1")+ elif thinking_type == 'format':+ # Groq Qwen/QwQ: reasoning_format parameterformat_val = thinking.format if thinking.format in ('parsed', 'hidden') else 'parsed'kwargs['reasoning_format'] = format_val- elif provider == 'cerebras':- # Cerebras: No official LangChain thinking support yet- # Passing through as model_kwargs if supported- kwargs['thinking_budget'] = thinking.budget+ elif thinking_type == 'none':+ logger.warning(f"[AI] Model '{model}' ({provider}) may not support thinking mode")# Resolve lazy-loaded model class (gemini)model_class = config.model_class or _get_google_genai_class()@@ -1315,12 +1323,6 @@ class AIService:flattened.get('systemMessage') orflattened.get('systemPrompt') or '')- # Max tokens - support camelCase from frontend- max_tokens = int(flattened.get('max_tokens') or- flattened.get('maxTokens') or 1000)-- temperature = float(flattened.get('temperature', 0.7))-if not api_key:raise ValueError("API key is required")@@ -1339,6 +1341,9 @@ class AIService:else:provider = self.detect_provider(model)+ # Resolve max_tokens and temperature via model registry+ max_tokens = _resolve_max_tokens(flattened, model, provider)+# Build thinking config from parametersthinking_config = Noneif flattened.get('thinkingEnabled'):@@ -1350,6 +1355,8 @@ class AIService:format=flattened.get('reasoningFormat', 'parsed'),)+ temperature = _resolve_temperature(flattened, model, provider, bool(thinking_config and thinking_config.enabled))+# Check for proxy URL (stored as {provider}_proxy credential)proxy_url = await self.auth.get_api_key(f"{provider}_proxy")@@ -1518,8 +1525,6 @@ class AIService:api_key = flattened.get('api_key') or flattened.get('apiKey')provider = parameters.get('provider', 'openai')model = parameters.get('model', '')- temperature = float(flattened.get('temperature', 0.7))- max_tokens = int(flattened.get('max_tokens') or flattened.get('maxTokens') or 1000)logger.debug(f"[LangGraph] Agent: {provider}/{model}, tools={len(tool_data) if tool_data else 0}")@@ -1535,6 +1540,9 @@ class AIService:if not api_key:raise ValueError("API key is required for AI Agent")+ # Resolve max_tokens and temperature via model registry+ max_tokens = _resolve_max_tokens(flattened, model, provider)+# Build thinking config from parametersthinking_config = Noneif flattened.get('thinkingEnabled'):@@ -1547,6 +1555,8 @@ class AIService:)logger.debug(f"[LangGraph] Thinking enabled: budget={thinking_config.budget}, effort={thinking_config.effort}")+ temperature = _resolve_temperature(flattened, model, provider, bool(thinking_config and thinking_config.enabled))+# Broadcast: Initializing modelawait broadcast_status("initializing", {"message": f"Initializing {provider} model...",@@ -2008,8 +2018,6 @@ class AIService:api_key = flattened.get('api_key') or flattened.get('apiKey')provider = parameters.get('provider', 'openai')model = parameters.get('model', '')- temperature = float(flattened.get('temperature', 0.7))- max_tokens = int(flattened.get('max_tokens') or flattened.get('maxTokens') or 1000)logger.debug(f"[ChatAgent] Provider: {provider}, Model: {model}")@@ -2025,6 +2033,9 @@ class AIService:if not api_key:raise ValueError("API key is required for Zeenie")+ # Resolve max_tokens and temperature via model registry+ max_tokens = _resolve_max_tokens(flattened, model, provider)+# Build thinking config from parametersthinking_config = Noneif flattened.get('thinkingEnabled'):@@ -2036,6 +2047,8 @@ class AIService:format=flattened.get('reasoningFormat', 'parsed'),)+ temperature = _resolve_temperature(flattened, model, provider, bool(thinking_config and thinking_config.enabled))+# Broadcast: Initializingawait broadcast_status("initializing", {"message": f"Initializing {provider} model...",diff --git a/server/services/compaction.py b/server/services/compaction.pyindex 5a3ffb43..476c8d3a 100644--- a/server/services/compaction.py+++ b/server/services/compaction.py@@ -3,6 +3,9 @@Anthropic: tool_runner with compaction_control parameterOpenAI: context_management with compact_thresholdOthers: Client-side summarization fallback++Threshold strategy: per-session custom_threshold > model-aware threshold > global default.+Model-aware threshold = 50% of model's context window (e.g., 100K for a 200K model)."""from pydantic import BaseModel@@ -18,6 +21,9 @@ if TYPE_CHECKING:logger = get_logger(__name__)+# Fraction of context window used as compaction threshold when no custom threshold is set+CONTEXT_THRESHOLD_RATIO = 0.5+def _extract_text_from_response(content) -> str:"""Extract text content from various LLM response formats.@@ -69,28 +75,45 @@ class CompactionService:"""Wire AI service for generating compaction summaries."""self._ai_service = ai_service- def anthropic_config(self, threshold: Optional[int] = None) -> Dict[str, Any]:+ def get_model_threshold(self, model: str, provider: str) -> int:+ """Compute compaction threshold based on model's context window.++ Returns 50% of the model's context length, floored to the global+ default if the registry has no data. Per-session custom_threshold+ (checked in track()) still takes priority over this value.+ """+ from services.model_registry import get_model_registry+ registry = get_model_registry()+ context_length = registry.get_context_length(model, provider)+ model_threshold = int(context_length * CONTEXT_THRESHOLD_RATIO)+ # Never go below the configured minimum (ge=10000 in Settings)+ return max(model_threshold, 10000)++ def anthropic_config(self, threshold: Optional[int] = None, model: str = "", provider: str = "anthropic") -> Dict[str, Any]:"""Anthropic SDK compaction_control for tool_runner."""+ effective = threshold or self.get_model_threshold(model, provider)return {"enabled": self._config.enabled,- "context_token_threshold": threshold or self._config.threshold+ "context_token_threshold": effective}- def anthropic_api_config(self, threshold: Optional[int] = None) -> Dict[str, Any]:+ def anthropic_api_config(self, threshold: Optional[int] = None, model: str = "", provider: str = "anthropic") -> Dict[str, Any]:"""Anthropic Messages API context_management config."""+ effective = threshold or self.get_model_threshold(model, provider)return {"betas": ["compact-2026-01-12"],"context_management": {"edits": [{"type": "compact_20260112",- "trigger": {"type": "input_tokens", "value": threshold or self._config.threshold}+ "trigger": {"type": "input_tokens", "value": effective}}]}}- def openai_config(self, threshold: Optional[int] = None) -> Dict[str, Any]:+ def openai_config(self, threshold: Optional[int] = None, model: str = "", provider: str = "openai") -> Dict[str, Any]:"""OpenAI context_management config."""- return {"context_management": {"compact_threshold": threshold or self._config.threshold}}+ effective = threshold or self.get_model_threshold(model, provider)+ return {"context_management": {"compact_threshold": effective}}async def track(self, session_id: str, node_id: str, provider: str, model: str, usage: Dict[str, int]) -> Dict[str, Any]:"""Track token usage and cost, return compaction status."""@@ -133,7 +156,13 @@ class CompactionService:"cumulative_total_cost": new_total_cost})- threshold = state.get("custom_threshold") or self._config.threshold+ # Priority: per-session custom > model-aware > global default+ custom = state.get("custom_threshold")+ if custom:+ threshold = custom+ else:+ threshold = self.get_model_threshold(model, provider)+return {"total": new_total,"total_cost": new_total_cost,@@ -156,10 +185,26 @@ class CompactionService:"compaction_count": state["compaction_count"] + 1})- async def stats(self, session_id: str) -> Dict[str, Any]:- """Get session statistics."""+ async def stats(self, session_id: str, model: str = "", provider: str = "") -> Dict[str, Any]:+ """Get session statistics.++ When model/provider are provided, the threshold reflects the model's+ context window. Otherwise falls back to the global default.+ """state = await self._db.get_or_create_session_token_state(session_id)- return {"session_id": session_id, "total": state["cumulative_total"], "threshold": state.get("custom_threshold") or self._config.threshold, "count": state.get("compaction_count", 0)}+ custom = state.get("custom_threshold")+ if custom:+ threshold = custom+ elif model and provider:+ threshold = self.get_model_threshold(model, provider)+ else:+ threshold = self._config.threshold+ return {+ "session_id": session_id,+ "total": state["cumulative_total"],+ "threshold": threshold,+ "count": state.get("compaction_count", 0),+ }async def configure(self, session_id: str, threshold: Optional[int] = None, enabled: Optional[bool] = None) -> bool:"""Configure session settings."""@@ -218,9 +263,14 @@ CONVERSATION:Provide a concise but complete summary."""+ # Use a reasonable summary size: min(4096, model's max output)+ from services.model_registry import get_model_registry+ model_max = get_model_registry().get_max_output_tokens(model, provider)+ summary_tokens = min(4096, model_max)+llm = self._ai_service.create_model(provider=provider, api_key=api_key, model=model,- temperature=0.3, max_tokens=2000+ temperature=0.3, max_tokens=summary_tokens)from langchain_core.messages import HumanMessagediff --git a/server/services/handlers/telegram.py b/server/services/handlers/telegram.pyindex a669ee83..8b1aadf2 100644--- a/server/services/handlers/telegram.py+++ b/server/services/handlers/telegram.py@@ -54,17 +54,31 @@ async def handle_telegram_send("timestamp": datetime.now().isoformat()}- # Extract parameters- chat_id = parameters.get("chat_id")- if not chat_id:- return {- "success": False,- "node_id": node_id,- "node_type": node_type,- "error": "chat_id is required",- "execution_time": time.time() - start_time,- "timestamp": datetime.now().isoformat()- }+ # Resolve recipient based on type (mirrors WhatsApp self-mode pattern)+ recipient_type = parameters.get("recipient_type", "self")++ if recipient_type == "self":+ chat_id = service.owner_chat_id+ if not chat_id:+ return {+ "success": False,+ "node_id": node_id,+ "node_type": node_type,+ "error": "Bot owner not detected. Send any message to your bot first.",+ "execution_time": time.time() - start_time,+ "timestamp": datetime.now().isoformat()+ }+ else:+ chat_id = parameters.get("chat_id")+ if not chat_id:+ return {+ "success": False,+ "node_id": node_id,+ "node_type": node_type,+ "error": "chat_id is required",+ "execution_time": time.time() - start_time,+ "timestamp": datetime.now().isoformat()+ }message_type = parameters.get("message_type", "text")parse_mode = parameters.get("parse_mode") or Nonediff --git a/server/services/model_registry.py b/server/services/model_registry.pynew file mode 100644index 00000000..a76dca86--- /dev/null+++ b/server/services/model_registry.py@@ -0,0 +1,500 @@+"""Model Parameter Management Service.++Centralized registry for model metadata (max_output_tokens, context_length,+temperature constraints, thinking capabilities). Fetches from OpenRouter's+public API and caches to JSON. Falls back to llm_defaults.json offline.+"""++import json+import re+from dataclasses import dataclass, asdict, field+from datetime import datetime, timezone, timedelta+from pathlib import Path+from typing import Dict, Any, Optional, List, Tuple++import httpx++from core.logging import get_logger++logger = get_logger(__name__)++# =============================================================================+# CONSTANTS+# =============================================================================++OPENROUTER_MODELS_URL = "https://openrouter.ai/api/v1/models"+CACHE_FILE = Path(__file__).parent.parent / "config" / "model_registry.json"+CACHE_MAX_AGE = timedelta(hours=24)++# Provider normalization: OpenRouter provider -> MachinaOs provider+PROVIDER_MAP = {+ "google": "gemini",+ "meta-llama": "meta",+ "x-ai": "xai",+}++# Known reasoning models (always temperature=1, no user override)+REASONING_MODEL_PATTERNS = re.compile(+ r"^(o1|o3|o4|o3-mini|o4-mini)(-|$)", re.IGNORECASE+)++# Known thinking model patterns -> thinking_type+THINKING_PATTERNS: List[Tuple[str, str, str]] = [+ # (provider, model_pattern, thinking_type)+ ("anthropic", r"claude-(opus|sonnet|haiku)-(4|5|6)", "budget"),+ ("anthropic", r"claude-3[\.\-]5", "budget"),+ ("openai", r"^(o1|o3|o4)", "effort"),+ ("openai", r"^gpt-5", "effort"), # GPT-5 hybrid reasoning+ ("gemini", r"gemini-(2\.5|3)", "budget"),+ ("groq", r"qwen3", "format"),+ ("cerebras", r"qwen", "format"),+]++# Default temperature ranges per provider+DEFAULT_TEMP_RANGES = {+ "openai": (0.0, 2.0),+ "anthropic": (0.0, 1.0),+ "gemini": (0.0, 2.0),+ "groq": (0.0, 2.0),+ "cerebras": (0.0, 1.5),+ "openrouter": (0.0, 2.0),+}+++# =============================================================================+# DATA STRUCTURES+# =============================================================================++@dataclass+class ModelInfo:+ """Comprehensive model metadata."""+ id: str # "openai/gpt-5.2"+ name: str # "OpenAI: GPT-5.2"+ provider: str # "openai" (MachinaOs-normalized)+ local_id: str # "gpt-5.2"+ context_length: int = 128000+ max_output_tokens: int = 4096+ input_price_per_mtok: float = 0.0+ output_price_per_mtok: float = 0.0+ supports_thinking: bool = False+ thinking_type: str = "none" # "budget" | "effort" | "format" | "none"+ temperature_range: tuple = (0.0, 2.0)+ is_reasoning_model: bool = False+ supported_parameters: list = field(default_factory=list)++ def to_dict(self) -> dict:+ d = asdict(self)+ d["temperature_range"] = list(self.temperature_range)+ return d++ @classmethod+ def from_dict(cls, data: dict) -> "ModelInfo":+ data = dict(data)+ if "temperature_range" in data and isinstance(data["temperature_range"], list):+ data["temperature_range"] = tuple(data["temperature_range"])+ if "supported_parameters" not in data:+ data["supported_parameters"] = []+ return cls(**data)+++# =============================================================================+# SERVICE+# =============================================================================++class ModelRegistryService:+ """Centralized model parameter management.++ Singleton service that fetches model metadata from OpenRouter's public API+ and caches to a JSON file. Provides lookup methods for max_output_tokens,+ temperature constraints, thinking capabilities, etc.+ """++ def __init__(self):+ self._models: Dict[str, ModelInfo] = {} # keyed by "provider/local_id"+ self._cache_timestamp: Optional[datetime] = None+ self._llm_defaults: Dict[str, Any] = {}++ # -------------------------------------------------------------------------+ # Lifecycle+ # -------------------------------------------------------------------------++ def startup(self) -> None:+ """Synchronously load from cache file and llm_defaults.json."""+ self._load_llm_defaults()+ self._load_cache()+ count = len(self._models)+ if count > 0:+ logger.info(f"Model registry loaded {count} models from cache")+ else:+ logger.info("Model registry: no cache found, will fetch on first refresh")++ def is_stale(self) -> bool:+ """Check if cache is older than 24h or missing."""+ if not self._cache_timestamp:+ return True+ age = datetime.now(timezone.utc) - self._cache_timestamp+ return age > CACHE_MAX_AGE++ async def refresh(self) -> int:+ """Fetch model data from OpenRouter public API and update cache.++ Returns the number of models loaded.+ """+ logger.info("Model registry: fetching from OpenRouter...")+ async with httpx.AsyncClient(timeout=30.0) as client:+ response = await client.get(OPENROUTER_MODELS_URL)+ response.raise_for_status()+ data = response.json()++ raw_models = data.get("data", [])+ if not raw_models:+ logger.warning("Model registry: OpenRouter returned no models")+ return 0++ new_models: Dict[str, ModelInfo] = {}+ for raw in raw_models:+ info = self._parse_openrouter_model(raw)+ if info:+ key = f"{info.provider}/{info.local_id}"+ new_models[key] = info++ self._models = new_models+ self._cache_timestamp = datetime.now(timezone.utc)+ self._save_cache()++ logger.info(f"Model registry: loaded {len(new_models)} models from OpenRouter")+ return len(new_models)++ # -------------------------------------------------------------------------+ # Core lookups+ # -------------------------------------------------------------------------++ def get_model_info(self, model: str, provider: str) -> Optional[ModelInfo]:+ """Look up model info with waterfall strategy.++ 1. Exact match on provider/model+ 2. Prefix match for versioned IDs+ 3. Cross-provider check (for OpenRouter models)+ 4. Returns None if not found+ """+ # Strip [FREE] prefix if present+ model = model.replace("[FREE] ", "")++ # 1. Exact match+ key = f"{provider}/{model}"+ if key in self._models:+ return self._models[key]++ # 2. Prefix match (e.g., claude-3-5-sonnet-20241022 -> claude-3-5-sonnet)+ for stored_key, info in self._models.items():+ if info.provider == provider and model.startswith(info.local_id):+ return info++ # 3. Cross-provider: try OpenRouter key format directly+ for stored_key, info in self._models.items():+ if info.local_id == model:+ return info++ # 4. For OpenRouter, try stripping provider prefix from model+ if provider == "openrouter" and "/" in model:+ _, local = model.split("/", 1)+ for stored_key, info in self._models.items():+ if info.local_id == local:+ return info++ return None++ def get_max_output_tokens(self, model: str, provider: str) -> int:+ """Get max output tokens: registry -> llm_defaults -> 4096."""+ info = self.get_model_info(model, provider)+ if info and info.max_output_tokens > 0:+ return info.max_output_tokens++ # Fallback to llm_defaults.json+ return self._get_default_max_output_tokens(provider, model)++ def get_context_length(self, model: str, provider: str) -> int:+ """Get context window length: registry -> llm_defaults -> 128000."""+ info = self.get_model_info(model, provider)+ if info and info.context_length > 0:+ return info.context_length+ return self._get_default_context_length(provider, model)++ def get_temperature_range(self, model: str, provider: str) -> Tuple[float, float]:+ """Get valid temperature range for a model."""+ info = self.get_model_info(model, provider)+ if info:+ return info.temperature_range++ # Fallback to llm_defaults or provider default+ providers = self._llm_defaults.get("providers", {})+ prov_cfg = providers.get(provider, {})+ temp_range = prov_cfg.get("temperature_range")+ if temp_range and isinstance(temp_range, list) and len(temp_range) == 2:+ return tuple(temp_range)++ return DEFAULT_TEMP_RANGES.get(provider, (0.0, 2.0))++ def is_reasoning_model(self, model: str, provider: str) -> bool:+ """Check if model is a reasoning model (always temp=1)."""+ info = self.get_model_info(model, provider)+ if info:+ return info.is_reasoning_model++ # Fallback: check llm_defaults reasoning_models list+ providers = self._llm_defaults.get("providers", {})+ reasoning_list = providers.get(provider, {}).get("reasoning_models", [])+ for pattern in reasoning_list:+ if model.startswith(pattern):+ return True++ # Fallback: pattern match+ return bool(REASONING_MODEL_PATTERNS.match(model))++ def supports_thinking(self, model: str, provider: str) -> bool:+ """Check if model supports thinking/reasoning mode."""+ info = self.get_model_info(model, provider)+ if info:+ return info.supports_thinking+ return self.get_thinking_type(model, provider) != "none"++ def get_thinking_type(self, model: str, provider: str) -> str:+ """Get thinking type: 'budget', 'effort', 'format', or 'none'."""+ info = self.get_model_info(model, provider)+ if info:+ return info.thinking_type++ # Fallback: check llm_defaults+ providers = self._llm_defaults.get("providers", {})+ prov_cfg = providers.get(provider, {})++ # Check if model matches thinking_models patterns+ thinking_models = prov_cfg.get("thinking_models", [])+ for pattern in thinking_models:+ if model.startswith(pattern) or pattern in model:+ return prov_cfg.get("thinking_type", "none")++ # Check reasoning_models (these use effort type)+ reasoning_list = prov_cfg.get("reasoning_models", [])+ for pattern in reasoning_list:+ if model.startswith(pattern):+ return "effort"++ # Provider-level thinking_type (e.g., anthropic always uses budget)+ if prov_cfg.get("thinking_type") and prov_cfg.get("thinking_type") != "none":+ # Only apply if it's a modern enough model+ return self._infer_thinking_type(model, provider)++ return "none"++ def get_model_constraints(self, model: str, provider: str) -> dict:+ """Get all constraints for a model in one call (for frontend)."""+ info = self.get_model_info(model, provider)+ if info:+ return {+ "found": True,+ "model": model,+ "provider": provider,+ "max_output_tokens": info.max_output_tokens,+ "context_length": info.context_length,+ "temperature_range": list(info.temperature_range),+ "supports_thinking": info.supports_thinking,+ "thinking_type": info.thinking_type,+ "is_reasoning_model": info.is_reasoning_model,+ }++ # Build from fallbacks+ return {+ "found": False,+ "model": model,+ "provider": provider,+ "max_output_tokens": self.get_max_output_tokens(model, provider),+ "context_length": self.get_context_length(model, provider),+ "temperature_range": list(self.get_temperature_range(model, provider)),+ "supports_thinking": self.supports_thinking(model, provider),+ "thinking_type": self.get_thinking_type(model, provider),+ "is_reasoning_model": self.is_reasoning_model(model, provider),+ }++ # -------------------------------------------------------------------------+ # OpenRouter parsing+ # -------------------------------------------------------------------------++ def _parse_openrouter_model(self, raw: dict) -> Optional[ModelInfo]:+ """Parse a single model entry from OpenRouter API response."""+ model_id = raw.get("id", "")+ if not model_id or "/" not in model_id:+ return None++ or_provider, local_id = model_id.split("/", 1)+ provider = PROVIDER_MAP.get(or_provider, or_provider)++ # Extract top_provider data+ top_provider = raw.get("top_provider", {}) or {}+ max_completion = top_provider.get("max_completion_tokens") or 0++ # Extract pricing (string values in dollars per token)+ pricing = raw.get("pricing", {}) or {}+ try:+ input_price = float(pricing.get("prompt", "0")) * 1_000_000+ output_price = float(pricing.get("completion", "0")) * 1_000_000+ except (ValueError, TypeError):+ input_price = 0.0+ output_price = 0.0++ # Detect thinking support from supported_parameters+ supported_params = raw.get("supported_parameters", []) or []+ has_reasoning = "include_reasoning" in supported_params or "reasoning" in supported_params++ # Determine thinking type+ thinking_type = self._infer_thinking_type(local_id, provider)++ # If API says reasoning is supported but we didn't detect a type, use provider default+ if has_reasoning and thinking_type == "none":+ if provider == "anthropic":+ thinking_type = "budget"+ elif provider == "openai":+ thinking_type = "effort"+ elif provider in ("groq", "cerebras"):+ thinking_type = "format"++ # Determine if this is a reasoning model (always temp=1)+ is_reasoning = bool(REASONING_MODEL_PATTERNS.match(local_id))++ # Temperature range+ if is_reasoning:+ temp_range = (1.0, 1.0)+ else:+ temp_range = DEFAULT_TEMP_RANGES.get(provider, (0.0, 2.0))++ return ModelInfo(+ id=model_id,+ name=raw.get("name", model_id),+ provider=provider,+ local_id=local_id,+ context_length=raw.get("context_length", 0) or 0,+ max_output_tokens=max_completion,+ input_price_per_mtok=round(input_price, 4),+ output_price_per_mtok=round(output_price, 4),+ supports_thinking=thinking_type != "none",+ thinking_type=thinking_type,+ temperature_range=temp_range,+ is_reasoning_model=is_reasoning,+ supported_parameters=supported_params,+ )++ def _infer_thinking_type(self, model: str, provider: str) -> str:+ """Infer thinking type from known model patterns."""+ for pat_provider, pat_regex, pat_type in THINKING_PATTERNS:+ if provider == pat_provider and re.search(pat_regex, model, re.IGNORECASE):+ return pat_type+ return "none"++ # -------------------------------------------------------------------------+ # Cache management+ # -------------------------------------------------------------------------++ def _load_cache(self) -> None:+ """Load models from cache file."""+ if not CACHE_FILE.exists():+ return++ try:+ with open(CACHE_FILE, encoding="utf-8") as f:+ data = json.load(f)++ ts_str = data.get("_generated")+ if ts_str:+ self._cache_timestamp = datetime.fromisoformat(ts_str)++ raw_models = data.get("models", {})+ for key, model_data in raw_models.items():+ try:+ self._models[key] = ModelInfo.from_dict(model_data)+ except (TypeError, KeyError) as e:+ logger.debug(f"Skipping invalid model cache entry {key}: {e}")++ except Exception as e:+ logger.warning(f"Failed to load model registry cache: {e}")++ def _save_cache(self) -> None:+ """Save models to cache file."""+ try:+ cache_data = {+ "_generated": self._cache_timestamp.isoformat() if self._cache_timestamp else None,+ "_source": "openrouter",+ "_model_count": len(self._models),+ "models": {+ key: info.to_dict() for key, info in self._models.items()+ }+ }++ CACHE_FILE.parent.mkdir(parents=True, exist_ok=True)+ with open(CACHE_FILE, "w", encoding="utf-8") as f:+ json.dump(cache_data, f, indent=2, default=str)++ logger.debug(f"Model registry cache saved: {len(self._models)} models")++ except Exception as e:+ logger.warning(f"Failed to save model registry cache: {e}")++ # -------------------------------------------------------------------------+ # LLM defaults fallback+ # -------------------------------------------------------------------------++ def _load_llm_defaults(self) -> None:+ """Load llm_defaults.json for offline fallback."""+ defaults_path = Path(__file__).parent.parent / "config" / "llm_defaults.json"+ try:+ with open(defaults_path, encoding="utf-8") as f:+ self._llm_defaults = json.load(f)+ except Exception as e:+ logger.warning(f"Could not load llm_defaults.json: {e}")+ self._llm_defaults = {"providers": {}}++ def _get_default_max_output_tokens(self, provider: str, model: str) -> int:+ """Fallback: get max output tokens from llm_defaults.json."""+ providers = self._llm_defaults.get("providers", {})+ token_map = providers.get(provider, {}).get("max_output_tokens", {})++ # Exact match+ if model in token_map:+ return token_map[model]++ # Prefix match+ for key, val in token_map.items():+ if key != "_default" and isinstance(val, int) and model.startswith(key):+ return val++ return token_map.get("_default", 4096)++ def _get_default_context_length(self, provider: str, model: str) -> int:+ """Fallback: get context length from llm_defaults.json."""+ providers = self._llm_defaults.get("providers", {})+ ctx_map = providers.get(provider, {}).get("context_length", {})++ # Exact match+ if model in ctx_map:+ return ctx_map[model]++ # Prefix match+ for key, val in ctx_map.items():+ if key != "_default" and isinstance(val, int) and model.startswith(key):+ return val++ return ctx_map.get("_default", 128000)+++# =============================================================================+# SINGLETON+# =============================================================================++_service: Optional[ModelRegistryService] = None+++def get_model_registry() -> ModelRegistryService:+ """Get or create the model registry singleton."""+ global _service+ if _service is None:+ _service = ModelRegistryService()+ return _servicediff --git a/server/services/status_broadcaster.py b/server/services/status_broadcaster.pyindex 6f1b91d5..f217d13d 100644--- a/server/services/status_broadcaster.py+++ b/server/services/status_broadcaster.py@@ -440,7 +440,8 @@ class StatusBroadcaster:connected: bool,bot_id: Optional[int] = None,bot_username: Optional[str] = None,- bot_name: Optional[str] = None+ bot_name: Optional[str] = None,+ owner_chat_id: Optional[int] = None):"""Update Telegram bot connection status and broadcast."""import time@@ -449,6 +450,7 @@ class StatusBroadcaster:"bot_id": bot_id,"bot_username": bot_username,"bot_name": bot_name,+ "owner_chat_id": owner_chat_id,"timestamp": time.time()}diff --git a/server/services/telegram_service.py b/server/services/telegram_service.pyindex 11d12de0..3a92c1a9 100644--- a/server/services/telegram_service.py+++ b/server/services/telegram_service.py@@ -29,6 +29,7 @@ class TelegramService:self._connected: bool = Falseself._polling_task: Optional[asyncio.Task] = Noneself._bot_info: Dict[str, Any] = {}+ self._owner_chat_id: Optional[int] = None@classmethoddef get_instance(cls) -> "TelegramService":@@ -49,6 +50,16 @@ class TelegramService:"""Check if bot is connected and polling."""return self._connected and self._application is not None+ @property+ def owner_chat_id(self) -> Optional[int]:+ """Get the bot owner's chat ID (auto-captured from first private message)."""+ return self._owner_chat_id++ async def set_owner(self, chat_id: int):+ """Set owner chat_id (used to restore from credentials on reconnect)."""+ self._owner_chat_id = chat_id+ logger.info(f"[Telegram] Owner chat_id restored: {chat_id}")+def get_status(self) -> Dict[str, Any]:"""Get current connection status."""return {@@ -57,6 +68,7 @@ class TelegramService:"bot_username": self._bot_info.get("username"),"bot_name": self._bot_info.get("first_name"),"polling_active": self._polling_task is not None and not self._polling_task.done(),+ "owner_chat_id": self._owner_chat_id,}async def connect(self, token: str) -> Dict[str, Any]:@@ -158,6 +170,7 @@ class TelegramService:self._token = Noneself._connected = Falseself._polling_task = None+ self._owner_chat_id = None# Broadcast disconnected statusawait self._broadcast_status()@@ -198,6 +211,20 @@ class TelegramService:returnmsg = update.message++ # Auto-capture bot owner from first private message+ if self._owner_chat_id is None and msg.chat.type == "private" and msg.from_user:+ self._owner_chat_id = msg.from_user.id+ logger.info(f"[Telegram] Owner detected: @{msg.from_user.username} (ID: {msg.from_user.id})")+ # Persist to credentials DB so it survives reconnects+ try:+ from services.auth import get_auth_service+ auth = get_auth_service()+ await auth.store_api_key("telegram_owner_chat_id", str(msg.from_user.id))+ except Exception as persist_err:+ logger.warning(f"[Telegram] Failed to persist owner chat_id: {persist_err}")+ await self._broadcast_status()+event_data = self._format_message(msg)logger.debug(f"[Telegram] Message received: {event_data.get('content_type')} from {event_data.get('from_username', event_data.get('from_id'))}")@@ -292,6 +319,7 @@ class TelegramService:bot_id=self._bot_info.get("id"),bot_username=self._bot_info.get("username"),bot_name=self._bot_info.get("first_name"),+ owner_chat_id=self._owner_chat_id,)except Exception as e:logger.warning(f"[Telegram] Status broadcast failed: {e}")