Skip to content

Self-improving AI Employees for everyone

feat: Add WhatsApp newsletter/channel support and proxy service docs

trohitg0b2b1a07932f
diff --cc CLAUDE.md
index 3efae3b9,5132f4f4..bea49afa
--- a/CLAUDE.md
+++ b/CLAUDE.md
@@@ -18,6 -18,6 +18,7 @@@ This is a React Flow-based workflow aut
| **[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 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 |
++| **[Proxy Service](./docs-internal/proxy_service.md)** | Residential proxy provider management with template-based URL formatting, health scoring, and transparent HTTP node injection |
| **[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 |
| **[Execution Engine Design](./docs-internal/DESIGN.md)** | Architecture patterns, design standards, and implementation details for the workflow execution engine |
@@@ -373,8 -368,8 +374,8 @@@ 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 + 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)
+- **92 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 + 3 Proxy)
- - **WebSocket-First Architecture**: WebSocket as primary frontend-backend communication (119 message handlers)
+ - **WebSocket-First Architecture**: WebSocket as primary frontend-backend communication (120 message handlers)
- **Recent optimizations**: REST APIs replaced with WebSocket, AI endpoints migrated to Python, Android automation integrated
## Architecture Refactoring
@@@ -1371,8 -1347,8 +1379,8 @@@ All scripts in `scripts/` are cross-pla
See **[Scripts Reference](./docs-internal/SCRIPTS.md)** for full documentation.
## Current Status
-✅ **INodeProperties System**: Fully implemented with 89 functional node components
+✅ **INodeProperties System**: Fully implemented with 92 functional node components
- ✅ **WebSocket-First Architecture**: 119 message handlers replacing REST APIs
+ ✅ **WebSocket-First Architecture**: 120 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
diff --cc docs-internal/proxy_service.md
index 00000000,00000000..1b5c018f
new file mode 100644
--- /dev/null
+++ b/docs-internal/proxy_service.md
@@@ -1,0 -1,0 +1,295 @@@
++# Proxy Service
++
++Residential proxy provider management with template-based URL formatting, health-scored auto-selection, geo-targeting, sticky sessions, routing rules, and transparent injection on HTTP nodes.
++
++## Architecture
++
++```
++ ┌──────────────────────────────────────────┐
++ │ ProxyService (singleton) │
++ │ │
++ proxy_config ────►│ _providers: Dict[name, _ProviderRuntime]│
++ (AI tool) │ _routing_rules: List[RoutingRule] │
++ │ _daily_spend_usd: float │
++ │ │
++ │ get_proxy_url(url, params) -> str|None │
++ │ report_result(name, ProxyResult) │
++ └──────────┬───────────────────────────────┘
++ │
++ ┌────────────────┼────────────────┐
++ ▼ ▼ ▼
++ TemplateProxyProvider AuthService Database
++ (format_proxy_url) (credentials) (providers, rules)
++```
++
++### Request Flow
++
++```
++httpRequest(useProxy=true)
++ │
++ ▼
++_get_proxy_url_if_enabled()
++ │
++ ▼
++ProxyService.get_proxy_url(url, parameters)
++ │
++ ├── Match routing rule by domain (fnmatch glob)
++ ├── Pick provider: explicit > rule preferred > best score
++ ├── Load credentials from AuthService
++ ├── Build GeoTarget from params/rule/default
++ ├── Determine SessionType (rotating/sticky)
++ │
++ ▼
++TemplateProxyProvider.format_proxy_url()
++ │
++ ▼
++"http://user-country-us:[email protected]:7777"
++ │
++ ▼
++httpx.AsyncClient(proxy=proxy_url)
++```
++
++## Key Files
++
++| File | Description |
++|------|-------------|
++| `server/services/proxy/__init__.py` | Exports `get_proxy_service`, `init_proxy_service`, `ProxyService` |
++| `server/services/proxy/service.py` | ProxyService singleton - provider selection, URL generation, health scoring |
++| `server/services/proxy/providers.py` | TemplateProxyProvider - formats proxy URLs from JSON template config |
++| `server/services/proxy/models.py` | Pydantic models: ProviderConfig, RoutingRule, SessionType, GeoTarget, ProxyResult, ProviderStats |
++| `server/services/proxy/exceptions.py` | Custom exceptions: NoHealthyProviderError, BudgetExceededError, ProxyConfigError, ProviderError |
++| `server/services/handlers/proxy.py` | Node handlers: handle_proxy_request, handle_proxy_status, handle_proxy_config |
++| `server/services/handlers/http.py` | `_get_proxy_url_if_enabled()` shared helper for transparent proxy injection |
++| `server/services/handlers/tools.py` | `_execute_proxy_config()`, `_execute_proxy_request()`, `_execute_proxy_status()` AI tool handlers |
++| `server/services/handlers/document.py` | httpScraper handler with proxy injection |
++| `server/core/database.py` | CRUD: get/save/delete proxy providers and routing rules |
++| `server/models/database.py` | SQLModel tables: ProxyProvider, ProxyRoutingRule |
++| `server/core/container.py` | DI: `proxy_service()` factory |
++| `server/main.py` | Startup: `proxy_svc.startup()` after database init |
++| `server/constants.py` | `PROXY_NODE_TYPES` frozenset |
++| `client/src/nodeDefinitions/proxyNodes.ts` | 3 node definitions + shared `PROXY_PARAMETERS` array |
++| `server/skills/web_agent/proxy-config-skill/SKILL.md` | AI skill: proxy provider setup and management |
++| `server/skills/web_agent/http-request-skill/SKILL.md` | AI skill: HTTP requests with `useProxy: true` |
++
++## Proxy Nodes (3)
++
++### proxyRequest
++Direct proxy HTTP request with full controls. Parameters: method, url, headers, body, timeout, proxyProvider, proxyCountry, sessionType (rotating/sticky), stickyDuration, maxRetries, followRedirects. Includes retry/failover loop and result reporting to health scorer.
++
++### proxyConfig
++Dual-purpose node (workflow + AI tool). Operations:
++
++| Operation | Required Params | Description |
++|-----------|----------------|-------------|
++| `list_providers` | none | List all configured providers |
++| `add_provider` | name, gateway_host, gateway_port, url_template | Add provider with JSON template |
++| `update_provider` | name + fields | Update existing provider |
++| `remove_provider` | name | Delete provider |
++| `set_credentials` | name, username, password | Store proxy credentials via AuthService |
++| `test_provider` | name | Test via httpbin.org/ip |
++| `get_stats` | none | Usage and health statistics |
++| `add_routing_rule` | domain_pattern | Route domains to specific providers |
++| `list_routing_rules` | none | List all routing rules |
++| `remove_routing_rule` | rule_id | Delete routing rule |
++
++### proxyStatus
++View provider health stats. Optionally filter by provider name.
++
++## Transparent Proxy on HTTP Nodes
++
++The `httpRequest` and `httpScraper` nodes support `useProxy: true`. When set, `_get_proxy_url_if_enabled()` calls `ProxyService.get_proxy_url()` and passes the result to `httpx.AsyncClient(proxy=...)`. If proxy lookup fails, the request proceeds without proxy (graceful degradation).
++
++The AI Agent's `http_request` tool schema includes `useProxy: bool`. The LLM only sets the flag -- the proxy service handles provider selection, geo-targeting, and session type from its own configuration.
++
++```python
++# In server/services/ai.py
++class HttpRequestSchema(BaseModel):
++ url: str
++ method: str = "GET"
++ body: Optional[Dict[str, Any]] = None
++ useProxy: bool = False # LLM sets this, proxy service handles the rest
++```
++
++## Template System
++
++### url_template JSON
++
++Each provider has a JSON `url_template` that controls how credentials and parameters are encoded into the proxy URL. This eliminates provider-specific code -- one class (`TemplateProxyProvider`) handles all providers.
++
++**Fields:**
++| Field | Description | Example |
++|-------|-------------|---------|
++| `param_field` | Where params go: `"username"`, `"password"`, or `"none"` | `"username"` |
++| `username_prefix` | Template for base username. `{username}` replaced with actual. | `"{username}"` |
++| `username_param_separator` | Separator between username and first param | `"-"` |
++| `param_separator` | Separator between params | `"-"` |
++| `param_keys` | Map of param names to format strings. `{v}` = value. | `{"country": "country-{v}"}` |
++| `country_case` | `"lower"` or `"upper"` for country codes | `"lower"` |
++| `city_separator` | Replaces spaces in city/state names | `"_"` |
++
++### Example Templates
++
++**Smartproxy/Decodo** (username-based, dash separator):
++```json
++{
++ "param_field": "username",
++ "username_prefix": "{username}",
++ "username_param_separator": "-",
++ "param_separator": "-",
++ "param_keys": {
++ "country": "country-{v}",
++ "city": "city-{v}",
++ "state": "state-{v}",
++ "session_id": "session-{v}",
++ "session_duration": "sessTime-{v}"
++ },
++ "country_case": "lower",
++ "city_separator": "_"
++}
++```
++
++Result: `http://myuser-country-us-session-abc123:[email protected]:7777`
++
++**IPRoyal** (password-based):
++```json
++{
++ "param_field": "password",
++ "username_prefix": "{username}",
++ "param_separator": "_",
++ "param_keys": {
++ "country": "country-{v}",
++ "city": "city-{v}",
++ "session_id": "session-{v}"
++ },
++ "country_case": "lower",
++ "city_separator": "_"
++}
++```
++
++Result: `http://myuser:[email protected]:12321`
++
++## Health Scoring
++
++Providers are ranked by a composite score (0.0 - 1.0):
++
++| Weight | Factor | Formula |
++|--------|--------|---------|
++| 0.4 | Success rate | `successes / total_requests` |
++| 0.2 | Latency | `1.0 - (avg_ms / 10000)` |
++| 0.3 | Cost | `1.0 - (cost_per_gb / 10.0)` |
++| 0.1 | Freshness | `min(1.0, sample_count / 10)` |
++
++A provider is considered unhealthy when its success rate drops below 30%. The last 100 results are kept per provider.
++
++## Provider Selection Priority
++
++1. **Explicit**: `parameters.proxyProvider` specified by user/node
++2. **Routing rule**: Domain matches a rule with `preferred_providers`
++3. **Best score**: Auto-select highest-scoring healthy provider
++
++Fallback: if preferred providers from a routing rule are all unhealthy, falls back to best available provider.
++
++## Routing Rules
++
++Domain-based rules using `fnmatch` glob patterns:
++
++```python
++RoutingRule(
++ domain_pattern="*.linkedin.com",
++ preferred_providers=["smartproxy", "brightdata"],
++ required_country="US",
++ session_type=SessionType.STICKY,
++ sticky_duration_seconds=300,
++ max_retries=3,
++ failover=True,
++ min_success_rate=0.7,
++ priority=0, # lower = evaluated first
++)
++```
++
++Rules are matched first-match against the target URL's hostname.
++
++## Credential Storage
++
++Proxy credentials are stored via AuthService using the pattern `proxy_{name}_username` and `proxy_{name}_password`:
++
++```python
++# In _execute_proxy_config set_credentials operation:
++await auth_service.store_api_key(f"proxy_{name}_username", username, [])
++await auth_service.store_api_key(f"proxy_{name}_password", password, [])
++```
++
++On startup, `ProxyService._load_providers()` reads credentials from AuthService for each provider.
++
++## Database Tables
++
++### ProxyProvider
++
++| Column | Type | Description |
++|--------|------|-------------|
++| id | Integer (PK) | Auto-increment |
++| name | String (unique) | Provider identifier |
++| enabled | Boolean | Active flag |
++| priority | Integer | Lower = preferred |
++| cost_per_gb | Float | USD per GB |
++| gateway_host | String | Proxy gateway hostname |
++| gateway_port | Integer | Proxy gateway port |
++| url_template | Text (JSON) | Template config for URL formatting |
++| geo_coverage | Text (JSON) | Supported country codes |
++| sticky_support | Boolean | Supports sticky sessions |
++| max_sticky_seconds | Integer | Max sticky duration |
++| created_at | DateTime | Creation timestamp |
++
++### ProxyRoutingRule
++
++| Column | Type | Description |
++|--------|------|-------------|
++| id | Integer (PK) | Auto-increment |
++| domain_pattern | String | fnmatch glob pattern |
++| preferred_providers | Text (JSON) | Provider name array |
++| required_country | String | ISO country code |
++| session_type | String | "rotating" or "sticky" |
++| sticky_duration_seconds | Integer | Sticky session duration |
++| max_retries | Integer | Retry attempts |
++| failover | Boolean | Try other providers on failure |
++| min_success_rate | Float | Health threshold |
++| priority | Integer | Rule evaluation order |
++
++## Configuration
++
++Environment variables in `server/.env`:
++
++```bash
++PROXY_ENABLED=true # Enable proxy service
++PROXY_DEFAULT_COUNTRY= # Default country code (empty = no default)
++PROXY_BUDGET_DAILY_USD= # Daily spend limit (empty = unlimited)
++```
++
++## Service Lifecycle
++
++```python
++# In server/main.py startup
++proxy_svc = container.proxy_service() # Creates singleton via init_proxy_service()
++await proxy_svc.startup() # Loads providers from DB + credentials from AuthService
++
++# In server/main.py shutdown
++await proxy_svc.shutdown() # Clears in-memory state
++```
++
++The proxy service always initializes regardless of `PROXY_ENABLED`. Providers are managed dynamically by the LLM via the `proxy_config` tool. If no providers are configured, `get_proxy_url()` raises `NoHealthyProviderError`.
++
++## Adding a New Provider (AI Workflow)
++
++The LLM uses the `proxy_config` tool in 3 steps:
++
++1. **Add provider** with gateway host, port, and url_template JSON
++2. **Set credentials** with username and password
++3. **Test provider** via httpbin.org/ip health check
++
++If the provider uses an unfamiliar URL format, the LLM can use `python_code` to reverse-engineer the template from the provider's documentation examples.
++
++## Cost Tracking
++
++Proxy usage is tracked via the pricing service. When a proxied request completes, `_track_proxy_usage()` in `handlers/proxy.py` calculates cost based on bytes transferred and the provider's `cost_per_gb` from `server/config/pricing.json`. The cost is persisted to the `APIUsageMetric` table.
++
++The daily spend is also tracked in-memory on `ProxyService._daily_spend_usd` and checked against `PROXY_BUDGET_DAILY_USD` before each request.
diff --cc server/config/model_registry.json
index f676ffdb,f676ffdb..4a70c317
--- a/server/config/model_registry.json
+++ b/server/config/model_registry.json
@@@ -1,8 -1,8 +1,38 @@@
{
-- "_generated": "2026-02-26T17:40:58.078412+00:00",
++ "_generated": "2026-02-28T07:14:28.108966+00:00",
"_source": "openrouter",
"_model_count": 343,
"models": {
++ "bytedance-seed/seed-2.0-mini": {
++ "id": "bytedance-seed/seed-2.0-mini",
++ "name": "ByteDance Seed: Seed-2.0-Mini",
++ "provider": "bytedance-seed",
++ "local_id": "seed-2.0-mini",
++ "context_length": 262144,
++ "max_output_tokens": 131072,
++ "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",
++ "reasoning",
++ "response_format",
++ "stop",
++ "structured_outputs",
++ "temperature",
++ "tool_choice",
++ "tools",
++ "top_p"
++ ]
++ },
"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)",
@@@ -16,7 -16,7 +46,7 @@@
"thinking_type": "budget",
"temperature_range": [
0.0,
-- 1.0
++ 2.0
],
"is_reasoning_model": false,
"supported_parameters": [
@@@ -199,7 -199,7 +229,7 @@@
"thinking_type": "budget",
"temperature_range": [
0.0,
-- 1.0
++ 2.0
],
"is_reasoning_model": false,
"supported_parameters": [
@@@ -280,7 -280,7 +310,7 @@@
"thinking_type": "budget",
"temperature_range": [
0.0,
-- 1.0
++ 2.0
],
"is_reasoning_model": false,
"supported_parameters": [
@@@ -400,9 -400,9 +430,9 @@@
"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,
++ "max_output_tokens": 196608,
++ "input_price_per_mtok": 0.295,
++ "output_price_per_mtok": 1.2,
"supports_thinking": false,
"thinking_type": "none",
"temperature_range": [
@@@ -1188,7 -1188,7 +1218,7 @@@
"thinking_type": "budget",
"temperature_range": [
0.0,
-- 1.0
++ 2.0
],
"is_reasoning_model": false,
"supported_parameters": [
@@@ -1226,39 -1226,39 +1256,6 @@@
"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",
@@@ -1363,8 -1363,8 +1360,8 @@@
"max_output_tokens": 16384,
"input_price_per_mtok": 1.75,
"output_price_per_mtok": 14.0,
-- "supports_thinking": false,
-- "thinking_type": "none",
++ "supports_thinking": true,
++ "thinking_type": "effort",
"temperature_range": [
0.0,
2.0
@@@ -1853,6 -1853,6 +1850,8 @@@
"include_reasoning",
"max_tokens",
"reasoning",
++ "response_format",
++ "structured_outputs",
"temperature",
"top_p"
]
@@@ -2070,7 -2070,7 +2069,7 @@@
"thinking_type": "budget",
"temperature_range": [
0.0,
-- 1.0
++ 2.0
],
"is_reasoning_model": false,
"supported_parameters": [
@@@ -2129,7 -2129,7 +2128,7 @@@
"thinking_type": "budget",
"temperature_range": [
0.0,
-- 1.0
++ 2.0
],
"is_reasoning_model": false,
"supported_parameters": [
@@@ -2215,8 -2215,8 +2214,8 @@@
"max_output_tokens": 16384,
"input_price_per_mtok": 1.25,
"output_price_per_mtok": 10.0,
-- "supports_thinking": false,
-- "thinking_type": "none",
++ "supports_thinking": true,
++ "thinking_type": "effort",
"temperature_range": [
0.0,
2.0
@@@ -2989,7 -2989,7 +2988,7 @@@
"thinking_type": "budget",
"temperature_range": [
0.0,
-- 1.0
++ 2.0
],
"is_reasoning_model": false,
"supported_parameters": [
@@@ -3215,7 -3215,7 +3214,9 @@@
"supported_parameters": [
"frequency_penalty",
"include_reasoning",
++ "logit_bias",
"max_tokens",
++ "min_p",
"presence_penalty",
"reasoning",
"repetition_penalty",
@@@ -3296,7 -3296,7 +3297,7 @@@
"thinking_type": "budget",
"temperature_range": [
0.0,
-- 1.0
++ 2.0
],
"is_reasoning_model": false,
"supported_parameters": [
@@@ -4318,8 -4318,8 +4319,8 @@@
"max_output_tokens": 16384,
"input_price_per_mtok": 1.25,
"output_price_per_mtok": 10.0,
-- "supports_thinking": false,
-- "thinking_type": "none",
++ "supports_thinking": true,
++ "thinking_type": "effort",
"temperature_range": [
0.0,
2.0
@@@ -4991,7 -4991,7 +4992,7 @@@
"thinking_type": "budget",
"temperature_range": [
0.0,
-- 1.0
++ 2.0
],
"is_reasoning_model": false,
"supported_parameters": [
@@@ -5241,7 -5241,7 +5242,7 @@@
"thinking_type": "none",
"temperature_range": [
0.0,
-- 1.0
++ 2.0
],
"is_reasoning_model": false,
"supported_parameters": [
@@@ -5529,7 -5529,7 +5530,7 @@@
"thinking_type": "budget",
"temperature_range": [
0.0,
-- 1.0
++ 2.0
],
"is_reasoning_model": false,
"supported_parameters": [
@@@ -5559,7 -5559,7 +5560,7 @@@
"thinking_type": "budget",
"temperature_range": [
0.0,
-- 1.0
++ 2.0
],
"is_reasoning_model": false,
"supported_parameters": [
@@@ -5680,7 -5680,7 +5681,7 @@@
"thinking_type": "budget",
"temperature_range": [
0.0,
-- 1.0
++ 2.0
],
"is_reasoning_model": false,
"supported_parameters": [
@@@ -5704,8 -5704,8 +5705,8 @@@
"local_id": "deepseek-r1-0528",
"context_length": 163840,
"max_output_tokens": 65536,
-- "input_price_per_mtok": 0.4,
-- "output_price_per_mtok": 1.75,
++ "input_price_per_mtok": 0.45,
++ "output_price_per_mtok": 2.15,
"supports_thinking": false,
"thinking_type": "none",
"temperature_range": [
@@@ -5804,7 -5804,7 +5805,7 @@@
"thinking_type": "none",
"temperature_range": [
0.0,
-- 1.0
++ 2.0
],
"is_reasoning_model": false,
"supported_parameters": [
@@@ -5831,7 -5831,7 +5832,7 @@@
"thinking_type": "none",
"temperature_range": [
0.0,
-- 1.0
++ 2.0
],
"is_reasoning_model": false,
"supported_parameters": [
@@@ -5890,7 -5890,7 +5891,7 @@@
"thinking_type": "budget",
"temperature_range": [
0.0,
-- 1.0
++ 2.0
],
"is_reasoning_model": false,
"supported_parameters": [
@@@ -6158,7 -6158,7 +6159,7 @@@
"name": "Qwen: Qwen3 8B",
"provider": "qwen",
"local_id": "qwen3-8b",
-- "context_length": 32000,
++ "context_length": 40960,
"max_output_tokens": 8192,
"input_price_per_mtok": 0.05,
"output_price_per_mtok": 0.4,
@@@ -6695,9 -6695,9 +6696,9 @@@
"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,
++ "max_output_tokens": 0,
++ "input_price_per_mtok": 0.2,
++ "output_price_per_mtok": 0.77,
"supports_thinking": false,
"thinking_type": "none",
"temperature_range": [
@@@ -6839,7 -6839,7 +6840,7 @@@
"thinking_type": "none",
"temperature_range": [
0.0,
-- 1.0
++ 2.0
],
"is_reasoning_model": false,
"supported_parameters": [
@@@ -6864,7 -6864,7 +6865,7 @@@
"thinking_type": "none",
"temperature_range": [
0.0,
-- 1.0
++ 2.0
],
"is_reasoning_model": false,
"supported_parameters": [
@@@ -6894,7 -6894,7 +6895,7 @@@
"thinking_type": "none",
"temperature_range": [
0.0,
-- 1.0
++ 2.0
],
"is_reasoning_model": false,
"supported_parameters": [
@@@ -6918,7 -6918,7 +6919,7 @@@
"thinking_type": "none",
"temperature_range": [
0.0,
-- 1.0
++ 2.0
],
"is_reasoning_model": false,
"supported_parameters": [
@@@ -7025,7 -7025,7 +7026,7 @@@
"thinking_type": "none",
"temperature_range": [
0.0,
-- 1.0
++ 2.0
],
"is_reasoning_model": false,
"supported_parameters": [
@@@ -7052,7 -7052,7 +7053,7 @@@
"thinking_type": "none",
"temperature_range": [
0.0,
-- 1.0
++ 2.0
],
"is_reasoning_model": false,
"supported_parameters": [
@@@ -7203,6 -7203,6 +7204,7 @@@
"supported_parameters": [
"frequency_penalty",
"include_reasoning",
++ "logprobs",
"max_tokens",
"presence_penalty",
"reasoning",
@@@ -7213,6 -7213,6 +7215,7 @@@
"tool_choice",
"tools",
"top_k",
++ "top_logprobs",
"top_p"
]
},
@@@ -7229,7 -7229,7 +7232,7 @@@
"thinking_type": "none",
"temperature_range": [
0.0,
-- 1.0
++ 2.0
],
"is_reasoning_model": false,
"supported_parameters": [
@@@ -7394,7 -7394,7 +7397,7 @@@
"thinking_type": "none",
"temperature_range": [
0.0,
-- 1.0
++ 2.0
],
"is_reasoning_model": false,
"supported_parameters": [
@@@ -7720,6 -7720,6 +7723,7 @@@
"supported_parameters": [
"frequency_penalty",
"include_reasoning",
++ "logprobs",
"max_tokens",
"presence_penalty",
"reasoning",
@@@ -7730,6 -7730,6 +7734,7 @@@
"structured_outputs",
"temperature",
"top_k",
++ "top_logprobs",
"top_p"
]
},
@@@ -7958,6 -7958,6 +7963,7 @@@
"is_reasoning_model": false,
"supported_parameters": [
"frequency_penalty",
++ "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
@@@ -7968,6 -7968,6 +7974,7 @@@
"structured_outputs",
"temperature",
"top_k",
++ "top_logprobs",
"top_p"
]
},
@@@ -8367,6 -8367,6 +8374,7 @@@
"is_reasoning_model": false,
"supported_parameters": [
"frequency_penalty",
++ "logprobs",
"max_tokens",
"presence_penalty",
"response_format",
@@@ -8375,6 -8375,6 +8383,7 @@@
"temperature",
"tool_choice",
"tools",
++ "top_logprobs",
"top_p"
]
},
@@@ -8594,6 -8594,6 +8603,7 @@@
"supported_parameters": [
"frequency_penalty",
"logit_bias",
++ "logprobs",
"max_tokens",
"min_p",
"presence_penalty",
@@@ -8606,6 -8606,6 +8616,7 @@@
"tool_choice",
"tools",
"top_k",
++ "top_logprobs",
"top_p"
]
},
@@@ -9307,7 -9307,7 +9318,7 @@@
"thinking_type": "none",
"temperature_range": [
0.0,
-- 1.0
++ 2.0
],
"is_reasoning_model": false,
"supported_parameters": [
@@@ -9334,7 -9334,7 +9345,7 @@@
"thinking_type": "none",
"temperature_range": [
0.0,
-- 1.0
++ 2.0
],
"is_reasoning_model": false,
"supported_parameters": [