Catalogue LLM — Musubi

Version 2026.07.15 · servi signé sur /v1/latest · JSON brut sur /catalog.json

97modèles

5Frontier

22Large

57Medium

13Small

16plateformes

Palier	Rang	Score	Plateforme	Modèle	Contexte	req/j	tok/min	Capacités
Frontier	1	3999	google	`gemini-3.5-flash` Gemini 3.5 Flash	1048.6k	20	250k	toolsvision
Frontier	2	3998	cloudflare	`@cf/moonshotai/kimi-k2.6` Kimi K2.6 (CF)	262.1k	—	—	tools
Frontier	3	3997	nvidia	`moonshotai/kimi-k2.6` Kimi K2.6 (NV)	131.1k	—	—	tools
Frontier	4	3996	nvidia	`deepseek-ai/deepseek-v4-pro` DeepSeek V4 Pro (NV)	131.1k	—	—	tools
Frontier	5	3995	nvidia	`minimaxai/minimax-m2.7` MiniMax M2.7 (NV)	196.6k	—	—	tools
Large	1	2999	google	`gemma-4-31b-it` Gemma 4 31B IT	32.8k	1k	250k	vision
Large	2	2998	openrouter	`google/gemma-4-31b-it:free` Gemma 4 31B (free)	262.1k	200	—	—
Large	3	2997	ovh	`Qwen3.5-397B-A17B` Qwen3.5-397B-A17B	262.1k	—	—	—
Large	4	2996	cloudflare	`@cf/google/gemma-4-26b-a4b-it` Gemma 4 26B-A4B it (CF)	262.1k	—	—	—
Large	5	2995	google	`gemma-4-26b-a4b-it` Gemma 4 26B IT	32.8k	1k	250k	vision
Large	6	2994	openrouter	`google/gemma-4-26b-a4b-it:free` Gemma 4 26B-A4B (free)	262.1k	200	—	—
Large	7	2993	nvidia	`deepseek-ai/deepseek-v4-flash` DeepSeek V4 Flash (NV)	131.1k	—	—	tools
Large	8	2992	opencode	`deepseek-v4-flash-free` DeepSeek V4 Flash Free (OpenCode Zen)	131.1k	200	—	tools
Large	9	2991	opencode	`mimo-v2.5-free` MiMo-V2.5 Free (OpenCode Zen)	131.1k	200	—	—
Large	10	2990	openrouter	`qwen/qwen3-next-80b-a3b-instruct:free` Qwen3-Next 80B (free)	262.1k	200	—	tools
Large	11	2989	cerebras	`zai-glm-4.7` GLM-4.7 (Cerebras)	8.2k	—	30k	tools
Large	12	2988	ollama	`glm-4.7` GLM-4.7 (Ollama)	131.1k	—	—	tools
Large	13	2987	kilo	`nvidia/nemotron-3-super-120b-a12b:free` Nemotron 3 Super 120B (Kilo)	262.1k	—	—	tools
Large	14	2986	nvidia	`nvidia/nemotron-3-super-120b-a12b` Nemotron 3 Super 120B (NV)	262.1k	—	—	tools
Large	15	2985	openrouter	`nvidia/nemotron-3-super-120b-a12b:free` Nemotron 3 Super 120B (free)	1000k	200	—	tools
Large	16	2984	cerebras	`gpt-oss-120b` GPT-OSS 120B (Cerebras)	131.1k	—	30k	tools
Large	17	2983	cloudflare	`@cf/openai/gpt-oss-120b` GPT-OSS 120B (CF)	131.1k	—	—	tools
Large	18	2982	groq	`openai/gpt-oss-120b` GPT-OSS 120B (Groq)	131.1k	1k	8k	tools
Large	19	2981	ollama	`gpt-oss:120b` GPT-OSS 120B (Ollama)	131.1k	—	—	tools
Large	20	2980	openrouter	`openai/gpt-oss-120b:free` GPT-OSS 120B (free)	131.1k	1k	—	tools
Large	21	2979	ovh	`gpt-oss-120b` gpt-oss-120b	131.1k	—	—	—
Large	22	2978	google	`gemini-2.5-flash` Gemini 2.5 Flash	1048.6k	20	250k	toolsvision
Medium	1	1999	cloudflare	`@cf/zai-org/glm-4.7-flash` GLM-4.7 Flash (CF)	131.1k	—	—	tools
Medium	2	1998	zhipu	`glm-4.7-flash` GLM-4.7 Flash	131.1k	—	—	tools
Medium	3	1997	nvidia	`nvidia/nemotron-3-nano-30b-a3b` Nemotron 3 Nano 30B (NV)	262.1k	—	—	—
Medium	4	1996	openrouter	`nvidia/nemotron-3-nano-30b-a3b:free` Nemotron 3 Nano 30B (free)	262.1k	200	—	—
Medium	5	1995	ovh	`Qwen3-32B` Qwen3-32B	32.8k	—	—	—
Medium	6	1994	cohere	`command-a-03-2025` Command-A (03-2025)	131.1k	33	—	tools
Medium	7	1993	mistral	`mistral-medium-latest` Mistral Medium 3.5	131.1k	—	500k	tools
Medium	8	1992	cloudflare	`@cf/qwen/qwen3-30b-a3b-fp8` Qwen3 30B-A3B fp8 (CF)	131.1k	—	—	tools
Medium	9	1991	nvidia	`meta/llama-4-maverick-17b-128e-instruct` Llama 4 Maverick (NV)	131.1k	—	—	toolsvision
Medium	10	1990	groq	`openai/gpt-oss-20b` GPT-OSS 20B (Groq)	131.1k	1k	8k	tools
Medium	11	1989	ollama	`gpt-oss:20b` GPT-OSS 20B (Ollama)	131.1k	—	—	tools
Medium	12	1988	openrouter	`openai/gpt-oss-20b:free` GPT-OSS 20B (free)	131.1k	200	—	tools
Medium	13	1987	ovh	`gpt-oss-20b` gpt-oss-20b	131.1k	—	—	—
Medium	14	1986	cloudflare	`@cf/meta/llama-4-scout-17b-16e-instruct` Llama 4 Scout (CF)	131.1k	—	—	tools
Medium	15	1985	ovh	`Qwen2.5-VL-72B-Instruct` Qwen2.5-VL-72B-Instruct	32.8k	—	—	vision
Medium	16	1984	mistral	`mistral-large-latest` Mistral Large 3	262.1k	—	500k	tools
Medium	17	1983	mistral	`magistral-medium-latest` Magistral Medium	131.1k	—	500k	tools
Medium	18	1982	cloudflare	`@cf/deepseek-ai/deepseek-r1-distill-qwen-32b` DeepSeek R1 Distill Qwen 32B (CF)	131.1k	—	—	tools
Medium	19	1981	cloudflare	`@cf/ibm-granite/granite-4.0-h-micro` Granite 4.0 H Micro (CF)	131.1k	—	—	—
Medium	20	1980	cloudflare	`@cf/nvidia/nemotron-3-120b-a12b` Nemotron 3 120B (CF)	262.1k	—	—	—
Medium	21	1979	cohere	`command-a-reasoning-08-2025` Command A Reasoning (08-2025)	256k	33	—	tools
Medium	22	1978	github	`openai/gpt-4.1` GPT-4.1 (GitHub)	8k	50	—	toolsvision
Medium	23	1977	google	`gemini-2.5-flash-lite` Gemini 2.5 Flash-Lite	1048.6k	20	250k	toolsvision
Medium	24	1976	groq	`groq/compound` Compound (Groq)	131.1k	250	70k	vision
Medium	25	1975	groq	`groq/compound-mini` Compound Mini (Groq)	131.1k	250	70k	—
Medium	26	1974	groq	`openai/gpt-oss-safeguard-20b` GPT-OSS Safeguard 20B (Groq)	131.1k	1k	8k	tools
Medium	27	1973	kilo	`cohere/north-mini-code:free` Cohere: North Mini Code (free)	256k	—	—	—
Medium	28	1972	kilo	`nvidia/nemotron-3-nano-omni-30b-a3b-reasoning:free` NVIDIA: Nemotron 3 Nano Omni (free)	256k	—	—	—
Medium	29	1971	kilo	`nvidia/nemotron-3-ultra-550b-a55b:free` NVIDIA: Nemotron 3 Ultra (free)	1000k	—	—	—
Medium	30	1970	kilo	`poolside/laguna-m.1:free` Poolside Laguna M.1 (Kilo)	262.1k	—	—	—
Medium	31	1969	kilo	`poolside/laguna-xs-2.1:free` Poolside: Laguna XS 2.1 (free)	262.1k	—	—	—
Medium	32	1968	kilo	`stepfun/step-3.7-flash:free` StepFun Step 3.7 Flash (Kilo)	262.1k	—	—	—
Medium	33	1967	kilo	`tencent/hy3:free` Tencent: Hy3 (free)	262.1k	—	—	—
Medium	34	1966	llm7	`codestral-latest` Codestral (LLM7)	32k	—	—	tools
Medium	35	1965	mistral	`codestral-latest` Codestral	256k	—	500k	tools
Medium	36	1964	mistral	`devstral-latest` Devstral	262.1k	—	500k	tools
Medium	37	1963	nvidia	`mistralai/mistral-large-3-675b-instruct-2512` Mistral Large 3 675B (NV)	131.1k	—	—	tools
Medium	38	1962	ollama	`gemma4:31b` Gemma 4 31B (Ollama)	131.1k	—	—	—
Medium	39	1961	ollama	`qwen3-coder-next` Qwen3-Coder Next (Ollama)	262.1k	—	—	tools
Medium	40	1960	ollama	`qwen3-coder:480b` Qwen3-Coder 480B (Ollama)	262.1k	—	—	tools
Medium	41	1959	opencode	`big-pickle` Big Pickle (OpenCode Zen, stealth)	131.1k	200	—	—
Medium	42	1958	openrouter	`cognitivecomputations/dolphin-mistral-24b-venice-edition:free` Dolphin Mistral 24B Venice (free)	32.8k	200	—	—
Medium	43	1957	openrouter	`liquid/lfm-2.5-1.2b-instruct:free` Liquid LFM 2.5 1.2B (free)	32.8k	200	—	—
Medium	44	1956	openrouter	`liquid/lfm-2.5-1.2b-thinking:free` Liquid LFM 2.5 1.2B Thinking (free)	32.8k	200	—	—
Medium	45	1955	openrouter	`nousresearch/hermes-3-llama-3.1-405b:free` Hermes 3 405B (free)	131.1k	50	—	—
Medium	46	1954	openrouter	`nvidia/nemotron-3-nano-omni-30b-a3b-reasoning:free` Nemotron 3 Nano 30B Reasoning (free)	262.1k	200	—	—
Medium	47	1953	openrouter	`nvidia/nemotron-3-ultra-550b-a55b:free` Nemotron 3 Ultra 550B (free, slow)	1000k	200	—	tools
Medium	48	1952	openrouter	`nvidia/nemotron-nano-9b-v2:free` Nemotron Nano 9B v2 (free)	128k	200	—	—
Medium	49	1951	openrouter	`poolside/laguna-m.1:free` Poolside Laguna M.1 (free)	131.1k	200	—	—
Medium	50	1950	openrouter	`qwen/qwen3-coder:free` Qwen3 Coder 480B (free)	1048.6k	50	—	tools
Medium	51	1949	ovh	`Mistral-7B-Instruct-v0.3` Mistral-7B-Instruct-v0.3	65.5k	—	—	—
Medium	52	1948	ovh	`Mistral-Small-3.2-24B-Instruct-2506` Mistral-Small-3.2-24B-Instruct-2506	131.1k	—	—	—
Medium	53	1947	ovh	`Qwen3-Coder-30B-A3B-Instruct` Qwen3-Coder-30B-A3B-Instruct	262.1k	—	—	—
Medium	54	1946	ovh	`Qwen3.5-9B` Qwen3.5-9B	262.1k	—	—	—
Medium	55	1945	ovh	`Qwen3.6-27B` Qwen3.6-27B	262.1k	—	—	—
Medium	56	1944	pollinations	`openai-fast` GPT-OSS 20B (Pollinations)	131.1k	—	—	—
Medium	57	1943	zhipu	`glm-4.6v-flash` GLM-4.6V Flash	131.1k	—	—	toolsvision
Small	1	999	mistral	`mistral-small-latest` Mistral Small 4	262.1k	—	500k	tools
Small	2	998	cloudflare	`@cf/meta/llama-3.3-70b-instruct-fp8-fast` Llama 3.3 70B fp8-fast (CF)	24k	—	—	tools
Small	3	997	groq	`llama-3.3-70b-versatile` Llama 3.3 70B	131.1k	1k	12k	tools
Small	4	996	nvidia	`meta/llama-3.3-70b-instruct` Llama 3.3 70B (NV)	131.1k	—	—	tools
Small	5	995	openrouter	`meta-llama/llama-3.3-70b-instruct:free` Llama 3.3 70B (OR free)	131.1k	1k	—	tools
Small	6	994	ovh	`Meta-Llama-3_3-70B-Instruct` Meta-Llama-3_3-70B-Instruct	131.1k	—	—	—
Small	7	993	cohere	`command-r-plus-08-2024` Command R+ (08-2024)	131.1k	33	—	tools
Small	8	992	nvidia	`meta/llama-3.1-70b-instruct` Llama 3.1 70B (NV)	131.1k	—	—	tools
Small	9	991	mistral	`ministral-8b-latest` Ministral 3 8B	262.1k	—	500k	tools
Small	10	990	cohere	`command-r-08-2024` Command R (08-2024)	131.1k	33	—	tools
Small	11	989	ovh	`Mistral-Nemo-Instruct-2407` Mistral-Nemo-Instruct-2407	65.5k	—	—	—
Small	12	988	groq	`llama-3.1-8b-instant` Llama 3.1 8B Instant	131.1k	14.4k	6k	tools
Small	13	987	openrouter	`meta-llama/llama-3.2-3b-instruct:free` Llama 3.2 3B (free)	131.1k	200	—	—

Score = palier × 1000 − rang — la formule exacte du routeur (router.ts:402). Le palier domine ; le rang ne départage qu'à l'intérieur d'un palier.
Classement dérivé de données Epoch AI et LMArena (CC-BY) ; métadonnées de modèles via OpenRouter. Vitesse et fiabilité mesurées en live par le routeur.