Self-improving AI Employees for everyone
Python ai-agentsai-employeeai-employeesassistantautonomous-agentsco-employeescompanygrok-botinstinctloop-agentsmuseown-your-datapersonalself-hosted
chore(config): curate latest models from the 2026-09-11 registry
trohitgab3731db74cd
diff --git a/docs-internal/native_llm_sdk.md b/docs-internal/native_llm_sdk.mdindex b6c15558..44e37d53 100644--- a/docs-internal/native_llm_sdk.md+++ b/docs-internal/native_llm_sdk.md@@ -93,7 +93,7 @@ Source of truth for this list: `server/config/llm_defaults.json` (the `providers| Provider | Key Models | Context | Max Output | Thinking | Temp Range ||----------|-----------|---------|-----------|----------|------------|-| **OpenAI** | GPT-5.6 Sol/Terra/Luna (+ `-pro`; default `gpt-5.6-sol`), GPT-5.5/5.4 | 1.05M | 128K | effort | omitted for reasoning models |+| **OpenAI** | GPT-6 Astra (+ `-pro`), GPT-5.6 Sol/Terra/Luna (+ `-pro`; default `gpt-5.6-sol`), GPT-5.5/5.4 | 1.05M | 128K | effort | omitted for reasoning models || **OpenAI** | GPT-4.1 | ~1.05M | 32K | none | 0-2 || **Anthropic** | Claude Opus 5 (default `claude-opus-5`), Fable 5.1 / 5, Sonnet 5 | 1M | 128K | adaptive | omitted (`temperature` / `top_p` / `top_k` rejected — `sampling_params_removed`) || **Anthropic** | Claude Opus 4.8/4.7 | 1M | 128K | adaptive | omitted (`sampling_params_removed`) |@@ -101,7 +101,7 @@ Source of truth for this list: `server/config/llm_defaults.json` (the `providers| **Anthropic** | Claude Haiku 4.5 | 200K | 64K | budget | 0-1 || **Google** | Gemini 3.8-flash (default), 3.7/3.6/3.5-flash, 3.5-flash-lite, 3.1-pro-preview/flash-lite, 3-flash-preview, 2.5-pro/flash/flash-lite | 1M | 64K | budget (`thinking_level` on 3.x when set explicitly) | 0-2 || **xAI** | Grok 4.20/4.20-multi-agent, 4.5, 4.3, 3 | 131K-2M | 131K | model/provider dependent | 0-2 |-| **DeepSeek** | deepseek-v4-flash, deepseek-v4-pro (deepseek-chat/reasoner legacy) | 1M | 64K | thinking modes | 0-2 |+| **DeepSeek** | deepseek-v4-flash (default), deepseek-v4.1-flash, deepseek-v4-pro (deepseek-chat/reasoner legacy) | 1M | 64K | thinking modes | 0-2 || **Kimi** | kimi-k3 (default), kimi-k2.6, kimi-k2.5, kimi-k2.7-code | 1M (K3); 256K (K2) | 131K (K3); 32K/96K (K2) | K2 provider default explicitly disabled unless requested | K2 fixed 0.6; K3 0-1 || **Mistral** | mistral-large/medium/small-latest, codestral-latest | 256K | 131K | none | 0-1.5 || **Groq** | GPT-OSS-120b/20b, Qwen3-32b, legacy Llama 3.x tiers | 131K | 32K-131K | effort (GPT-OSS), format (Qwen3) | 0-2 |diff --git a/docs-internal/workflow-schema.md b/docs-internal/workflow-schema.mdindex 799cac4f..c0d08c9a 100644--- a/docs-internal/workflow-schema.md+++ b/docs-internal/workflow-schema.md@@ -91,7 +91,7 @@ A workflow JSON document contains:- `openrouterChatModel` - OpenRouter unified API (200+ models)- `groqChatModel` - Groq ultra-fast inference (Llama, Qwen3, GPT-OSS)- `cerebrasChatModel` - Cerebras custom AI hardware (Llama, Qwen)-- `deepseekChatModel` - DeepSeek V4 (deepseek-v4-flash / deepseek-v4-pro; chat/reasoner legacy aliases)+- `deepseekChatModel` - DeepSeek V4 (deepseek-v4-flash / deepseek-v4.1-flash / deepseek-v4-pro; chat/reasoner legacy aliases)- `kimiChatModel` - Moonshot Kimi K2.6 / K2.5 / K2.7-code- `mistralChatModel` - Mistral Large / Small / Codestral- `ollamaChatModel` - Local Ollama server (OpenAI-compat via `{provider}_proxy` URL)diff --git a/server/config/llm_defaults.json b/server/config/llm_defaults.jsonindex 129f9a3c..be803778 100644--- a/server/config/llm_defaults.json+++ b/server/config/llm_defaults.json@@ -1,6 +1,6 @@{- "version": "2026.09.03",- "last_updated": "2026-09-03",+ "version": "2026.09.11",+ "last_updated": "2026-09-11","_description": "LLM provider defaults. popular_models drives the ordered global selector and offline fallback independently of token constraints. POLICY (2026-07-23): popular_models lists only models with >= 1M-token context windows — providers whose models are all below 1M carry an empty list and contribute nothing to the global selector (their models stay reachable via the live per-provider fetch_models dropdown). base_url directly configures the nine OpenAI-compatible native providers; dedicated OpenAI, Anthropic, and Gemini providers use SDK defaults, while OpenRouter has its own gateway default. Runtime proxy credentials can override supported endpoints. Model specs are enriched by model_registry.json (OpenRouter cache).","agent": {@@ -16,7 +16,7 @@"openai": {"display_name": "OpenAI","default_model": "gpt-5.6-sol",- "popular_models": ["gpt-5.6-sol", "gpt-5.6-sol-pro", "gpt-5.6-terra", "gpt-5.6-terra-pro", "gpt-5.6-luna", "gpt-5.6-luna-pro", "gpt-5.5", "gpt-5.4", "gpt-4.1"],+ "popular_models": ["gpt-5.6-sol", "gpt-6-astra", "gpt-6-astra-pro", "gpt-5.6-sol-pro", "gpt-5.6-terra", "gpt-5.6-terra-pro", "gpt-5.6-luna", "gpt-5.6-luna-pro", "gpt-5.5", "gpt-5.4", "gpt-4.1"],"detection_patterns": ["gpt", "openai", "o1", "o3", "o4"],"base_url": "https://api.openai.com/v1","models_endpoint": "https://api.openai.com/v1/models",@@ -26,9 +26,11 @@"enabled": false,"_note": "Native image input is emitted only when enabled is true; unknown/absent providers get a text placeholder instead (a rejected image block would poison journaled history). Read by services/llm/media.py provider_supports_vision. False here until the OpenAI request encoder renders image blocks (Responses input_image in function_call_output; Chat Completions hoist-to-user-turn) — flipping it early would hydrate bytes the encoder silently drops. Anthropic ships the first encoder; text-only providers keep the visionAnalyze delegate tool."},- "_models_note": "GPT-5.6 family (sol/terra/luna) GA 2026-07-09. gpt-5.5 / gpt-5.4 remain GA as legacy tiers. Dropped: gpt-5.2 (new requests blocked 2026-06-30), o3 (API shutdown 2026-12-11), o4-mini (API shutdown 2026-10-23) — they fall back to _default constraints if still selected before shutdown.",+ "_models_note": "GPT-6 Astra (+ -pro) appear in the 2026-09-11 registry refresh ($10/$50 per MTok, 1.05M context, 128K max output, effort thinking) and are curated as the top tier; the default stays gpt-5.6-sol ($2/$10), the same price-over-flagship call as anthropic keeping claude-opus-5 over claude-fable-5-1. The rolling -latest aliases (gpt-astra-latest, gpt-sol-latest, gpt-terra-latest, gpt-luna-latest) are deliberately not curated: real model names only, like gemini. GPT-5.6 family (sol/terra/luna) GA 2026-07-09. gpt-5.5 / gpt-5.4 remain GA as legacy tiers. Dropped: gpt-5.2 (new requests blocked 2026-06-30), o3 (API shutdown 2026-12-11), o4-mini (API shutdown 2026-10-23) — they fall back to _default constraints if still selected before shutdown.","max_output_tokens": {"gpt-5.6-sol": 128000,+ "gpt-6-astra": 128000,+ "gpt-6-astra-pro": 128000,"gpt-5.6-sol-pro": 128000,"gpt-5.6-terra": 128000,"gpt-5.6-terra-pro": 128000,@@ -41,6 +43,8 @@},"context_length": {"gpt-5.6-sol": 1050000,+ "gpt-6-astra": 1050000,+ "gpt-6-astra-pro": 1050000,"gpt-5.6-sol-pro": 1050000,"gpt-5.6-terra": 1050000,"gpt-5.6-terra-pro": 1050000,@@ -52,7 +56,7 @@"_default": 128000},"reasoning_models": ["o4-mini", "o3"],- "thinking_models": ["gpt-5"],+ "thinking_models": ["gpt-5", "gpt-6"],"thinking_type": "effort","temperature_range": [0.0, 2.0]},@@ -170,7 +174,7 @@"openrouter": {"display_name": "OpenRouter","default_model": "anthropic/claude-sonnet-5",- "popular_models": ["anthropic/claude-sonnet-5", "openai/gpt-5.6-terra", "openai/gpt-5.6-sol", "google/gemini-3.5-flash", "moonshotai/kimi-k3", "x-ai/grok-4.20"],+ "popular_models": ["anthropic/claude-sonnet-5", "openai/gpt-5.6-terra", "openai/gpt-5.6-sol", "openai/gpt-6-astra", "google/gemini-3.5-flash", "moonshotai/kimi-k3", "x-ai/grok-4.20"],"detection_patterns": ["openrouter"],"base_url": "https://openrouter.ai/api/v1","models_endpoint": "https://openrouter.ai/api/v1/models",@@ -276,8 +280,9 @@"deepseek": {"display_name": "DeepSeek","default_model": "deepseek-v4-flash",- "popular_models": ["deepseek-v4-flash", "deepseek-v4-pro"],+ "popular_models": ["deepseek-v4-flash", "deepseek-v4.1-flash", "deepseek-v4-pro"],"_popular_note": "deepseek-chat / deepseek-reasoner dropped from the global list per the 1M-context policy (131072); reachable via live fetch_models.",+ "_models_note": "deepseek-v4.1-flash appears in the 2026-09-11 registry refresh ($0.15/$0.60 per MTok, 1M context; the router cache reports a 384K max output, curated at the same 65536 as v4-flash until DeepSeek documents the cap). Default stays deepseek-v4-flash ($0.085/$0.17). Dated snapshots (-0731 / -0813), the vision-exp preview and the -latest alias are deliberately not curated.","detection_patterns": ["deepseek"],"base_url": "https://api.deepseek.com","models_endpoint": "https://api.deepseek.com/models",@@ -286,6 +291,7 @@"supported_params": ["temperature", "max_tokens", "top_p", "frequency_penalty", "presence_penalty", "stop", "seed", "tools", "tool_choice", "response_format"],"max_output_tokens": {"deepseek-v4-flash": 65536,+ "deepseek-v4.1-flash": 65536,"deepseek-v4-pro": 65536,"deepseek-chat": 8192,"deepseek-reasoner": 65536,@@ -293,6 +299,7 @@},"context_length": {"deepseek-v4-flash": 1048576,+ "deepseek-v4.1-flash": 1048576,"deepseek-v4-pro": 1048576,"deepseek-chat": 131072,"deepseek-reasoner": 131072,diff --git a/server/config/model_registry.json b/server/config/model_registry.jsonindex 23e1815b..c17cdb5a 100644--- a/server/config/model_registry.json+++ b/server/config/model_registry.json@@ -1,8 +1,285 @@{- "_generated": "2026-09-09T07:22:39.821023+00:00",+ "_generated": "2026-09-11T20:22:56.766352+00:00","_source": "openrouter",- "_model_count": 431,+ "_model_count": 443,"models": {+ "openai/gpt-astra-latest": {+ "id": "~openai/gpt-astra-latest",+ "name": "OpenAI GPT Astra Latest",+ "provider": "openai",+ "local_id": "gpt-astra-latest",+ "context_length": 1050000,+ "max_output_tokens": 128000,+ "input_price_per_mtok": 10.0,+ "output_price_per_mtok": 50.0,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_completion_tokens",+ "max_tokens",+ "reasoning",+ "reasoning_effort",+ "response_format",+ "seed",+ "structured_outputs",+ "tool_choice",+ "tools"+ ]+ },+ "openai/gpt-sol-latest": {+ "id": "~openai/gpt-sol-latest",+ "name": "OpenAI GPT Sol Latest",+ "provider": "openai",+ "local_id": "gpt-sol-latest",+ "context_length": 1050000,+ "max_output_tokens": 128000,+ "input_price_per_mtok": 2.0,+ "output_price_per_mtok": 10.0,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_completion_tokens",+ "max_tokens",+ "reasoning",+ "reasoning_effort",+ "response_format",+ "seed",+ "structured_outputs",+ "tool_choice",+ "tools"+ ]+ },+ "openai/gpt-terra-latest": {+ "id": "~openai/gpt-terra-latest",+ "name": "OpenAI GPT Terra Latest",+ "provider": "openai",+ "local_id": "gpt-terra-latest",+ "context_length": 1050000,+ "max_output_tokens": 128000,+ "input_price_per_mtok": 2.0,+ "output_price_per_mtok": 12.0,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_completion_tokens",+ "max_tokens",+ "reasoning",+ "reasoning_effort",+ "response_format",+ "seed",+ "structured_outputs",+ "tool_choice",+ "tools"+ ]+ },+ "openai/gpt-luna-latest": {+ "id": "~openai/gpt-luna-latest",+ "name": "OpenAI GPT Luna Latest",+ "provider": "openai",+ "local_id": "gpt-luna-latest",+ "context_length": 1050000,+ "max_output_tokens": 128000,+ "input_price_per_mtok": 0.2,+ "output_price_per_mtok": 1.2,+ "supports_thinking": true,+ "thinking_type": "effort",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "max_completion_tokens",+ "max_tokens",+ "reasoning",+ "reasoning_effort",+ "response_format",+ "seed",+ "structured_outputs",+ "tool_choice",+ "tools"+ ]+ },+ "sakana/fugu-ultra-v2": {+ "id": "sakana/fugu-ultra-v2",+ "name": "Sakana: Fugu Ultra v2",+ "provider": "sakana",+ "local_id": "fugu-ultra-v2",+ "context_length": 1000000,+ "max_output_tokens": 128000,+ "input_price_per_mtok": 5.0,+ "output_price_per_mtok": 30.0,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "include_reasoning",+ "reasoning",+ "reasoning_effort",+ "structured_outputs",+ "tool_choice",+ "tools",+ "web_search_options"+ ]+ },+ "sakana/fugu-max": {+ "id": "sakana/fugu-max",+ "name": "Sakana: Fugu Max",+ "provider": "sakana",+ "local_id": "fugu-max",+ "context_length": 1000000,+ "max_output_tokens": 128000,+ "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": [+ "include_reasoning",+ "reasoning",+ "reasoning_effort",+ "structured_outputs",+ "tool_choice",+ "tools",+ "web_search_options"+ ]+ },+ "inclusionai/ling-3.0-flash-vl": {+ "id": "inclusionai/ling-3.0-flash-vl",+ "name": "inclusionAI: Ling 3.0 Flash VL",+ "provider": "inclusionai",+ "local_id": "ling-3.0-flash-vl",+ "context_length": 131072,+ "max_output_tokens": 32768,+ "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",+ "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"+ ]+ },+ "inclusionai/ling-3.0-flash-vl:free": {+ "id": "inclusionai/ling-3.0-flash-vl:free",+ "name": "inclusionAI: Ling 3.0 Flash VL (free)",+ "provider": "inclusionai",+ "local_id": "ling-3.0-flash-vl:free",+ "context_length": 262144,+ "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",+ "logprobs",+ "max_tokens",+ "presence_penalty",+ "reasoning",+ "repetition_penalty",+ "seed",+ "stop",+ "temperature",+ "tool_choice",+ "tools",+ "top_k",+ "top_logprobs",+ "top_p"+ ]+ },+ "deepseek/deepseek-v4.1-flash": {+ "id": "deepseek/deepseek-v4.1-flash",+ "name": "DeepSeek: DeepSeek V4.1 Flash",+ "provider": "deepseek",+ "local_id": "deepseek-v4.1-flash",+ "context_length": 1048576,+ "max_output_tokens": 384000,+ "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",+ "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"+ ]+ },"inception/mercury-2.5": {"id": "inception/mercury-2.5","name": "Inception: Mercury 2.5",@@ -605,7 +882,7 @@"provider": "z-ai","local_id": "glm-flash-latest","context_length": 1310720,- "max_output_tokens": 943718,+ "max_output_tokens": 131072,"input_price_per_mtok": 0.075,"output_price_per_mtok": 0.25,"supports_thinking": false,@@ -657,10 +934,13 @@"supported_parameters": ["frequency_penalty","include_reasoning",+ "logit_bias","logprobs","max_tokens",+ "min_p","presence_penalty","reasoning",+ "repetition_penalty","response_format","seed","stop",@@ -680,8 +960,8 @@"local_id": "glm-5.3-flash","context_length": 1310720,"max_output_tokens": 131072,- "input_price_per_mtok": 0.075,- "output_price_per_mtok": 0.25,+ "input_price_per_mtok": 0.15,+ "output_price_per_mtok": 0.5,"supports_thinking": false,"thinking_type": "none","temperature_range": [@@ -911,9 +1191,9 @@"provider": "z-ai","local_id": "glm-latest","context_length": 1310720,- "max_output_tokens": 128000,- "input_price_per_mtok": 1.113,- "output_price_per_mtok": 3.498,+ "max_output_tokens": 943718,+ "input_price_per_mtok": 0.8727,+ "output_price_per_mtok": 3.36,"supports_thinking": false,"thinking_type": "none","temperature_range": [@@ -1081,6 +1361,7 @@"temperature","tool_choice","tools",+ "top_a","top_k","top_logprobs","top_p"@@ -1318,7 +1599,7 @@"provider": "deepseek","local_id": "deepseek-v4-pro-0813","context_length": 1048576,- "max_output_tokens": 393216,+ "max_output_tokens": 384000,"input_price_per_mtok": 0.5795,"output_price_per_mtok": 1.7384,"supports_thinking": false,@@ -1559,8 +1840,8 @@"local_id": "solar-pro4","context_length": 524288,"max_output_tokens": 131072,- "input_price_per_mtok": 0.03,- "output_price_per_mtok": 0.12,+ "input_price_per_mtok": 0.09,+ "output_price_per_mtok": 0.36,"supports_thinking": false,"thinking_type": "none","temperature_range": [@@ -1903,7 +2184,6 @@"supported_parameters": ["frequency_penalty","include_reasoning",- "logprobs","max_tokens","presence_penalty","reasoning",@@ -1912,7 +2192,6 @@"stop","temperature","tools",- "top_logprobs","top_p"]},@@ -2254,7 +2533,7 @@"provider": "thinkingmachines","local_id": "inkling","context_length": 1048576,- "max_output_tokens": 471859,+ "max_output_tokens": 32768,"input_price_per_mtok": 1.0,"output_price_per_mtok": 4.05,"supports_thinking": false,@@ -2397,8 +2676,8 @@"local_id": "kimi-k3","context_length": 1048576,"max_output_tokens": 943718,- "input_price_per_mtok": 3.0,- "output_price_per_mtok": 15.0,+ "input_price_per_mtok": 1.89,+ "output_price_per_mtok": 9.48,"supports_thinking": false,"thinking_type": "none","temperature_range": [@@ -2999,8 +3278,8 @@"local_id": "hy3","context_length": 262144,"max_output_tokens": 128000,- "input_price_per_mtok": 0.132,- "output_price_per_mtok": 0.528,+ "input_price_per_mtok": 0.0825,+ "output_price_per_mtok": 0.33,"supports_thinking": false,"thinking_type": "none","temperature_range": [@@ -3287,9 +3566,9 @@"provider": "z-ai","local_id": "glm-5.2","context_length": 1048576,- "max_output_tokens": 131072,- "input_price_per_mtok": 0.966,- "output_price_per_mtok": 3.036,+ "max_output_tokens": 182476,+ "input_price_per_mtok": 0.6,+ "output_price_per_mtok": 2.0,"supports_thinking": false,"thinking_type": "none","temperature_range": [@@ -4331,8 +4610,8 @@"local_id": "kimi-latest","context_length": 1048576,"max_output_tokens": 943718,- "input_price_per_mtok": 2.4,- "output_price_per_mtok": 12.0,+ "input_price_per_mtok": 1.89,+ "output_price_per_mtok": 9.48,"supports_thinking": false,"thinking_type": "none","temperature_range": [@@ -4424,35 +4703,6 @@"verbosity"]},- "openai/gpt-latest": {- "id": "~openai/gpt-latest",- "name": "OpenAI GPT Latest",- "provider": "openai",- "local_id": "gpt-latest",- "context_length": 1050000,- "max_output_tokens": 128000,- "input_price_per_mtok": 2.0,- "output_price_per_mtok": 10.0,- "supports_thinking": true,- "thinking_type": "effort",- "temperature_range": [- 0.0,- 2.0- ],- "is_reasoning_model": false,- "supported_parameters": [- "include_reasoning",- "max_completion_tokens",- "max_tokens",- "reasoning",- "reasoning_effort",- "response_format",- "seed",- "structured_outputs",- "tool_choice",- "tools"- ]- },"qwen/qwen3.5-plus-20260420": {"id": "qwen/qwen3.5-plus-20260420","name": "Qwen: Qwen3.5 Plus 2026-04-20",@@ -4754,8 +5004,8 @@"local_id": "deepseek-v4-pro","context_length": 1048576,"max_output_tokens": 384000,- "input_price_per_mtok": 0.9553,- "output_price_per_mtok": 1.9105,+ "input_price_per_mtok": 0.9478,+ "output_price_per_mtok": 1.8956,"supports_thinking": false,"thinking_type": "none","temperature_range": [@@ -4794,8 +5044,8 @@"local_id": "deepseek-v4-flash","context_length": 1048576,"max_output_tokens": 384000,- "input_price_per_mtok": 0.0886,- "output_price_per_mtok": 0.1772,+ "input_price_per_mtok": 0.0851,+ "output_price_per_mtok": 0.1702,"supports_thinking": false,"thinking_type": "none","temperature_range": [@@ -5153,9 +5403,9 @@"provider": "gemini","local_id": "gemma-4-26b-a4b-it","context_length": 262144,- "max_output_tokens": 16384,- "input_price_per_mtok": 0.07,- "output_price_per_mtok": 0.34,+ "max_output_tokens": 32768,+ "input_price_per_mtok": 0.042,+ "output_price_per_mtok": 0.22,"supports_thinking": false,"thinking_type": "none","temperature_range": [@@ -5765,6 +6015,39 @@"top_p"]},+ "mistralai/mistral-small-2603:batch": {+ "id": "mistralai/mistral-small-2603:batch",+ "name": "Mistral: Mistral Small 4 (batch)",+ "provider": "mistralai",+ "local_id": "mistral-small-2603:batch",+ "context_length": 262144,+ "max_output_tokens": 209715,+ "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",+ "presence_penalty",+ "reasoning",+ "reasoning_effort",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },"z-ai/glm-5-turbo": {"id": "z-ai/glm-5-turbo","name": "Z.ai: GLM 5 Turbo",@@ -5813,6 +6096,7 @@"frequency_penalty","include_reasoning","logit_bias",+ "logprobs","max_tokens","min_p","presence_penalty",@@ -5822,10 +6106,12 @@"response_format","seed","stop",+ "structured_outputs","temperature","tool_choice","tools","top_k",+ "top_logprobs","top_p"]},@@ -6277,9 +6563,9 @@"provider": "qwen","local_id": "qwen3.5-122b-a10b","context_length": 262144,- "max_output_tokens": 81920,- "input_price_per_mtok": 0.29,- "output_price_per_mtok": 2.4,+ "max_output_tokens": 65536,+ "input_price_per_mtok": 0.26,+ "output_price_per_mtok": 2.08,"supports_thinking": false,"thinking_type": "none","temperature_range": [@@ -6684,10 +6970,8 @@"supported_parameters": ["frequency_penalty","include_reasoning",- "logit_bias","logprobs","max_tokens",- "min_p","presence_penalty","reasoning","repetition_penalty",@@ -7090,7 +7374,7 @@"name": "Z.ai: GLM 4.7 Flash","provider": "z-ai","local_id": "glm-4.7-flash",- "context_length": 202752,+ "context_length": 200000,"max_output_tokens": 117964,"input_price_per_mtok": 0.0605,"output_price_per_mtok": 0.4,@@ -7743,6 +8027,36 @@"top_p"]},+ "mistralai/ministral-8b-2512:batch": {+ "id": "mistralai/ministral-8b-2512:batch",+ "name": "Mistral: Ministral 3 8B 2512 (batch)",+ "provider": "mistralai",+ "local_id": "ministral-8b-2512:batch",+ "context_length": 262144,+ "max_output_tokens": 209715,+ "input_price_per_mtok": 0.075,+ "output_price_per_mtok": 0.075,+ "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",@@ -7803,6 +8117,36 @@"top_p"]},+ "mistralai/mistral-large-2512:batch": {+ "id": "mistralai/mistral-large-2512:batch",+ "name": "Mistral: Mistral Large 3 2512 (batch)",+ "provider": "mistralai",+ "local_id": "mistral-large-2512:batch",+ "context_length": 262144,+ "max_output_tokens": 209715,+ "input_price_per_mtok": 0.25,+ "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",+ "presence_penalty",+ "response_format",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },"deepseek/deepseek-v3.2": {"id": "deepseek/deepseek-v3.2","name": "DeepSeek: DeepSeek V3.2",@@ -8048,7 +8392,7 @@"provider": "moonshotai","local_id": "kimi-k2-thinking","context_length": 262144,- "max_output_tokens": 235929,+ "max_output_tokens": 100352,"input_price_per_mtok": 0.6,"output_price_per_mtok": 2.5,"supports_thinking": false,@@ -9209,35 +9553,6 @@"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": 117964,- "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",@@ -9335,6 +9650,36 @@"top_p"]},+ "mistralai/mistral-medium-3.1:batch": {+ "id": "mistralai/mistral-medium-3.1:batch",+ "name": "Mistral: Mistral Medium 3.1 (batch)",+ "provider": "mistralai",+ "local_id": "mistral-medium-3.1:batch",+ "context_length": 131072,+ "max_output_tokens": 104857,+ "input_price_per_mtok": 0.2,+ "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",+ "seed",+ "stop",+ "structured_outputs",+ "temperature",+ "tool_choice",+ "tools",+ "top_p"+ ]+ },"z-ai/glm-4.5v": {"id": "z-ai/glm-4.5v","name": "Z.ai: GLM 4.5V",@@ -9775,6 +10120,37 @@"top_p"]},+ "mistralai/codestral-2508:batch": {+ "id": "mistralai/codestral-2508:batch",+ "name": "Mistral: Codestral 2508 (batch)",+ "provider": "mistralai",+ "local_id": "codestral-2508:batch",+ "context_length": 256000,+ "max_output_tokens": 204800,+ "input_price_per_mtok": 0.15,+ "output_price_per_mtok": 0.45,+ "supports_thinking": false,+ "thinking_type": "none",+ "temperature_range": [+ 0.0,+ 2.0+ ],+ "is_reasoning_model": false,+ "supported_parameters": [+ "frequency_penalty",+ "max_tokens",+ "prediction",+ "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",@@ -9815,9 +10191,9 @@"provider": "qwen","local_id": "qwen3-30b-a3b-instruct-2507","context_length": 262144,- "max_output_tokens": 32000,- "input_price_per_mtok": 0.0481,- "output_price_per_mtok": 0.193,+ "max_output_tokens": 235929,+ "input_price_per_mtok": 0.09,+ "output_price_per_mtok": 0.3,"supports_thinking": false,"thinking_type": "none","temperature_range": [@@ -11247,8 +11623,8 @@"local_id": "deepseek-chat-v3-0324","context_length": 163840,"max_output_tokens": 147456,- "input_price_per_mtok": 0.29,- "output_price_per_mtok": 1.14,+ "input_price_per_mtok": 0.25,+ "output_price_per_mtok": 1.0,"supports_thinking": false,"thinking_type": "none","temperature_range": [@@ -11993,9 +12369,9 @@"provider": "deepseek","local_id": "deepseek-chat","context_length": 163840,- "max_output_tokens": 16384,- "input_price_per_mtok": 0.32,- "output_price_per_mtok": 0.89,+ "max_output_tokens": 16000,+ "input_price_per_mtok": 0.2574,+ "output_price_per_mtok": 1.0287,"supports_thinking": false,"thinking_type": "none","temperature_range": [diff --git a/server/services/model_registry.py b/server/services/model_registry.pyindex 589e14a3..d330e387 100644--- a/server/services/model_registry.py+++ b/server/services/model_registry.py@@ -42,7 +42,7 @@ THINKING_PATTERNS: List[Tuple[str, str, str]] = [("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+ ("openai", r"^gpt-(5|6)", "effort"), # GPT-5 / GPT-6 hybrid reasoning("gemini", r"gemini-(2\.5|3)", "budget"),("groq", r"qwen3", "format"),("cerebras", r"qwen", "budget"),