How many Together AI models are retired?
315 Together AI model records have status retired.
315 Together AI model records have status retired, sorted by shutdown date. The stored status is retired. The dataset records these models as already shut down. Every record on this page has a shutdown date. Each row links the model page and the replacement page. The official page is https://docs.together.ai/docs/deprecations. The model page records the day that page was checked. Last updated 2026-11-10, the dataset version (the newest changelog date). See all retired models. What retired means is on Retired.
| Model | Shutdown | Replacement |
|---|---|---|
| allenai-olmo-7b-twin-2t The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| allenai/OLMo-7B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| allenai/OLMo-7B-Instruct The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| Austism/chronos-hermes-13b The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| codellama/CodeLlama-13b-Instruct-hf The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| codellama/CodeLlama-13b-Python-hf The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| codellama/CodeLlama-34b-Instruct-hf The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| codellama/CodeLlama-34b-Python-hf The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| codellama/CodeLlama-70b-hf The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| codellama/CodeLlama-70b-Instruct-hf The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| codellama/CodeLlama-70b-Python-hf The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| codellama/CodeLlama-7b-Instruct-hf The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| codellama/CodeLlama-7b-Python-hf The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| cognitivecomputations/dolphin-2.5-mixtral-8x7b The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| deepseek-ai/deepseek-coder-33b-instruct The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| garage-bAInd/Platypus2-70B-instruct The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| google/gemma-2b The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| google/gemma-7b The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| google/gemma-7b-it The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| hazyresearch/M2-BERT-2k-Retrieval-Encoder-V1 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| lmsys/vicuna-13b-v1.5 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| lmsys/vicuna-7b-v1.5 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| meta-llama/Llama-2-13b-hf The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| meta-llama/Llama-2-70b-chat-hf The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| meta-llama/Llama-2-7b-chat-hf The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| meta-llama/Llama-2-7b-hf The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| meta-llama/Llama-3-8b-hf The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| meta-llama/Meta-Llama-3-70B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| microsoft/phi-2 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| mistralai/Mistral-7B-Instruct-v0.1-json The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| mistralai/Mistral-7B-Instruct-v0.1-tools The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| mistralai/Mixtral-8x22B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| Nexusflow/NexusRaven-V2-13B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| NousResearch/Nous-Capybara-7B-V1p9 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| NousResearch/Nous-Hermes-2-Mistral-7B-DPO The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| NousResearch/Nous-Hermes-2-Mixtral-8x7B-SFT The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| NousResearch/Nous-Hermes-llama-2-7b The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| NousResearch/Nous-Hermes-Llama2-13b The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| Open-Orca/Mistral-7B-OpenOrca The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| openchat/openchat-3.5-1210 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| Phind/Phind-CodeLlama-34B-v2 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| prompthero/openjourney The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| Qwen/Qwen1.5-0.5B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| Qwen/Qwen1.5-0.5B-Chat The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| Qwen/Qwen1.5-1.8B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| Qwen/Qwen1.5-1.8B-Chat The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| Qwen/Qwen1.5-14B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| Qwen/Qwen1.5-14B-Chat The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| Qwen/Qwen1.5-32B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| Qwen/Qwen1.5-32B-Chat The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| Qwen/Qwen1.5-4B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| Qwen/Qwen1.5-4B-Chat The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| Qwen/Qwen1.5-72B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| Qwen/Qwen1.5-7B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| Qwen/Qwen1.5-7B-Chat The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| runwayml/stable-diffusion-v1-5 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| SG161222/Realistic_Vision_V3.0_VAE The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| snorkelai/Snorkel-Mistral-PairRM-DPO The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| Snowflake/snowflake-arctic-instruct The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| stabilityai/stable-diffusion-2-1 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| teknium/OpenHermes-2-Mistral-7B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| teknium/OpenHermes-2p5-Mistral-7B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| togethercomputer-codellama-34b-instruct-json The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| togethercomputer-codellama-34b-instruct-tools The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| togethercomputer/alpaca-7b The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| togethercomputer/evo-1-131k-base The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| togethercomputer/evo-1-8k-base The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| togethercomputer/GPT-JT-Moderation-6B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| togethercomputer/LLaMA-2-7B-32K The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| togethercomputer/LLaMA-2-7B-32K-Instruct The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| togethercomputer/RedPajama-INCITE-7B-Base The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| togethercomputer/RedPajama-INCITE-7B-Chat The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| togethercomputer/RedPajama-INCITE-7B-Instruct The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| togethercomputer/RedPajama-INCITE-Base-3B-v1 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| togethercomputer/RedPajama-INCITE-Chat-3B-v1 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| togethercomputer/RedPajama-INCITE-Instruct-3B-v1 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| togethercomputer/StripedHyena-Hessian-7B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| Undi95/ReMM-SLERP-L2-13B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| Undi95/Toppy-M-7B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| wavymulder/Analog-Diffusion The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| WhereIsAI/UAE-Large-V1 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| WizardLM/WizardCoder-15B-V1.0 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| WizardLM/WizardCoder-Python-34B-V1.0 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| WizardLM/WizardLM-13B-V1.2 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| zero-one-ai/Yi-34B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| zero-one-ai/Yi-34B-Chat The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| zero-one-ai/Yi-6B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-08-22 source, checked 2026-11-12 | Not published |
| NousResearch/Hermes-3-Llama-3.1-405B-Turbo The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-10-07 source, checked 2026-11-12 | Not published |
| NousResearch/Nous-Hermes-2-Yi-34B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-10-07 source, checked 2026-11-12 | Not published |
| Qwen/Qwen1.5-110B-Chat The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-10-29 source, checked 2026-11-12 | Not published |
| Qwen/Qwen1.5-72B-Chat The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2024-10-29 source, checked 2026-11-12 | Not published |
| bert-base-uncased The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. The inference table also lists removal date 2024-08-22 for this id. | 2025-02-06 source, checked 2026-11-12 | Not published |
| sentence-transformers/msmarco-bert-base-dot-v5 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. The inference table also lists removal date 2024-08-22 for this id. | 2025-02-06 source, checked 2026-11-12 | Not published |
| Meta-Llama/Llama-Guard-7b The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-03-08 source, checked 2026-11-12 | Not published |
| meta-llama/Meta-Llama-3-70B-Instruct-Lite The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-03-11 source, checked 2026-11-12 | Not published |
| databricks-dbrx-instruct The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-03-13 source, checked 2026-11-12 | Not published |
| Qwen/QwQ-32B-Preview The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-03-25 source, checked 2026-11-12 | Not published |
| mistralai/Mistral-7B-v0.1 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. The fine-tuning table lists suggested replacement mistralai/Mixtral-8x7B-Instruct-v0.1 and removal date 2026-07-29. These models can no longer be used as a base model for a fine-tuning job. | 2025-03-27 source, checked 2026-11-12 | Not published |
| meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo-lora The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-04-04 source, checked 2026-11-12 | Not published |
| stabilityai/stable-diffusion-xl-base-1.0 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-04-14 source, checked 2026-11-12 | Not published |
| meta-llama-llama-3-2-1b-instruct-lora The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-04-24 source, checked 2026-11-12 | Not published |
| meta-llama-meta-llama-3-1-70b-instruct-turbo-lora The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-04-24 source, checked 2026-11-12 | Not published |
| meta-llama-meta-llama-3-1-8b-instruct-turbo-lora The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-04-24 source, checked 2026-11-12 | Not published |
| meta-llama-meta-llama-3-70b-instruct-turbo The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-04-24 source, checked 2026-11-12 | Not published |
| meta-llama/Llama-2-13b-chat-hf The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-04-24 source, checked 2026-11-12 | Not published |
| microsoft-wizardlm-2-8x22b The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-04-24 source, checked 2026-11-12 | Not published |
| upstage-solar-10-7b-instruct-v1 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-04-24 source, checked 2026-11-12 | Not published |
| meta-llama-llama-3-2-3b-instruct-turbo-lora The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-05-16 source, checked 2026-11-12 | Not published |
| meta-llama-meta-llama-3-8b-instruct-turbo The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-05-16 source, checked 2026-11-12 | Not published |
| google-gemma-2-9b-it The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-06-13 source, checked 2026-11-12 | Not published |
| google-gemma-2b-it The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-06-13 source, checked 2026-11-12 | Not published |
| gryphe-mythomax-l2-13b The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-06-13 source, checked 2026-11-12 | Not published |
| gryphe-mythomax-l2-13b-lite The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-06-13 source, checked 2026-11-12 | Not published |
| mistralai-mixtral-8x22b-instruct-v0-1 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-06-13 source, checked 2026-11-12 | Not published |
| mistralai-mixtral-8x7b-v0-1 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-06-13 source, checked 2026-11-12 | Not published |
| togethercomputer-m2-bert-80m-2k-retrieval The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-06-13 source, checked 2026-11-12 | Not published |
| togethercomputer-m2-bert-80m-8k-retrieval The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-06-13 source, checked 2026-11-12 | Not published |
| whereisai-uae-large-v1 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-06-13 source, checked 2026-11-12 | Not published |
| arcee-ai/arcee-blitz The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-07-25 source, checked 2026-11-12 | Not published |
| arcee-ai/caller The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-07-25 source, checked 2026-11-12 | Not published |
| arcee-ai/virtuoso-medium-v2 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-07-25 source, checked 2026-11-12 | Not published |
| black-forest-labs/FLUX.1-canny The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-08-28 source, checked 2026-11-12 | Not published |
| black-forest-labs/FLUX.1-depth The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-08-28 source, checked 2026-11-12 | Not published |
| black-forest-labs/FLUX.1-redux The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-08-28 source, checked 2026-11-12 | Not published |
| deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-2B and removal date 2026-07-29. These models can no longer be used as a base model for a fine-tuning job. | 2025-08-28 source, checked 2026-11-12 | Not published |
| google/gemma-2-27b-it The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-08-28 source, checked 2026-11-12 | Not published |
| meta-llama-llama-3-3-70b-instruct-lora The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-08-28 source, checked 2026-11-12 | Not published |
| meta-llama/Llama-3-8b-chat-hf The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-08-28 source, checked 2026-11-12 | Not published |
| meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-08-28 source, checked 2026-11-12 | Not published |
| meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-08-28 source, checked 2026-11-12 | Not published |
| meta-llama/Llama-Vision-Free The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-08-28 source, checked 2026-11-12 | Not published |
| meta-llama/Meta-Llama-3-8B-Instruct The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. The fine-tuning table lists suggested replacement meta-llama/Meta-Llama-3.1-8B-Instruct-Reference and removal date 2026-07-29. These models can no longer be used as a base model for a fine-tuning job. | 2025-08-28 source, checked 2026-11-12 | Not published |
| NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-08-28 source, checked 2026-11-12 | Not published |
| nvidia/Llama-3.1-Nemotron-70B-Instruct-HF The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-08-28 source, checked 2026-11-12 | Not published |
| perplexity-ai/r1-1776 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-08-28 source, checked 2026-11-12 | Not published |
| Qwen/Qwen2-72B-Instruct The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. The fine-tuning table lists suggested replacement meta-llama/Llama-3.3-70B-Instruct-Reference and removal date 2026-07-29. These models can no longer be used as a base model for a fine-tuning job. | 2025-08-28 source, checked 2026-11-12 | Not published |
| Qwen/Qwen2-VL-72B-Instruct The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-08-28 source, checked 2026-11-12 | Not published |
| Qwen/Qwen2.5-14B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-27B and removal date 2026-07-29. These models can no longer be used as a base model for a fine-tuning job. | 2025-08-28 source, checked 2026-11-12 | Not published |
| deepseek-ai/DeepSeek-R1-Distill-Llama-70B-free The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-11-13 source, checked 2026-11-12 | Not published |
| deepseek-ai/DeepSeek-R1-Distill-Qwen-14B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-27B and removal date 2026-07-29. These models can no longer be used as a base model for a fine-tuning job. | 2025-11-13 source, checked 2026-11-12 | Not published |
| meta-llama/Llama-3.3-70B-Instruct-Turbo-Free The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-11-13 source, checked 2026-11-12 | Not published |
| mistralai/Mistral-7B-Instruct-v0.1 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-11-13 source, checked 2026-11-12 | Not published |
| Qwen/Qwen2.5-Coder-32B-Instruct The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-11-13 source, checked 2026-11-12 | Not published |
| Qwen/QwQ-32B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-11-13 source, checked 2026-11-12 | Not published |
| arcee_ai/arcee-spotlight The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-11-17 source, checked 2026-11-12 | Not published |
| arcee-ai/coder-large The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-11-17 source, checked 2026-11-12 | Not published |
| arcee-ai/maestro-reasoning The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-11-17 source, checked 2026-11-12 | Not published |
| arcee-ai/virtuoso-large The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-11-17 source, checked 2026-11-12 | Not published |
| deepcogito/cogito-v2-preview-deepseek-671b The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-11-19 source, checked 2026-11-12 | Not published |
| black-forest-labs/FLUX.1-schnell-free The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-12-23 source, checked 2026-11-12 | Not published |
| deepseek-ai/DeepSeek-R1-Distill-Llama-70B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. The fine-tuning table lists suggested replacement meta-llama/Llama-3.3-70B-Instruct-Reference and removal date 2026-07-29. These models can no longer be used as a base model for a fine-tuning job. | 2025-12-23 source, checked 2026-11-12 | Not published |
| meta-llama/Meta-Llama-3-70B-Instruct-Turbo The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-12-23 source, checked 2026-11-12 | Not published |
| meta-llama/Meta-Llama-Guard-3-8B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2025-12-23 source, checked 2026-11-12 | Not published |
| Qwen/Qwen2.5-VL-72B-Instruct The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-01-05 source, checked 2026-11-12 | Not published |
| deepseek-ai/DeepSeek-R1-0528-tput The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-02-03 source, checked 2026-11-12 | Not published |
| BAAI/bge-large-en-v1.5 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-02-06 source, checked 2026-11-12 | Not published |
| meta-llama/Llama-4-Scout-17B-16E-Instruct The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: Yes. Models marked "Yes" can be deployed as on-demand dedicated endpoints, either under the listed ID or as the underlying base model of a serving variant (for example, a deprecated -FP8 or -Turbo ID). | 2026-02-06 source, checked 2026-11-12 | Not published |
| meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-02-06 source, checked 2026-11-12 | Not published |
| qwen-qwen2-5-14b-instruct-lora The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-02-06 source, checked 2026-11-12 | Not published |
| Qwen/Qwen2.5-72B-Instruct-Turbo The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-02-06 source, checked 2026-11-12 | Not published |
| Qwen/Qwen3-235B-A22B-fp8-tput The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-02-06 source, checked 2026-11-12 | Not published |
| Salesforce/Llama-Rank-V1 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-02-06 source, checked 2026-11-12 | Not published |
| togethercomputer/m2-bert-80M-32k-retrieval The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-02-06 source, checked 2026-11-12 | Not published |
| togethercomputer/Refuel-Llm-V2 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-02-06 source, checked 2026-11-12 | Not published |
| togethercomputer/Refuel-Llm-V2-Small The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-02-06 source, checked 2026-11-12 | Not published |
| Alibaba-NLP/gte-modernbert-base The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-02-25 source, checked 2026-11-12 | Not published |
| BAAI/bge-base-en-v1.5-vllm The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-02-25 source, checked 2026-11-12 | Not published |
| black-forest-labs/FLUX.1-dev The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-02-25 source, checked 2026-11-12 | Not published |
| black-forest-labs/FLUX.1-dev-lora The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-02-25 source, checked 2026-11-12 | Not published |
| black-forest-labs/FLUX.1-Kontext-dev The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-02-25 source, checked 2026-11-12 | Not published |
| marin-community/Marin-8B-Instruct The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-02-25 source, checked 2026-11-12 | Not published |
| meta-llama/Llama-3.2-3B-Instruct-Turbo-Classifier The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-02-25 source, checked 2026-11-12 | Not published |
| meta-llama/Llama-Guard-3-11B-Vision-Turbo The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-02-25 source, checked 2026-11-12 | Not published |
| meta-llama/LlamaGuard-2-8b The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-02-25 source, checked 2026-11-12 | Not published |
| meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-02-25 source, checked 2026-11-12 | Not published |
| mistralai/Ministral-3-14B-Instruct The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-02-25 source, checked 2026-11-12 | Not published |
| nvidia/Nvidia-Nemotron-Nano-9B-v2 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-02-25 source, checked 2026-11-12 | Not published |
| Qwen/Qwen3-Next-80B-A3B-Thinking The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-122B-A10B and removal date 2026-07-29. These models can no longer be used as a base model for a fine-tuning job. | 2026-02-25 source, checked 2026-11-12 | Not published |
| Qwen/Qwen3-VL-32B-Instruct The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. The fine-tuning table lists removal date 2026-07-29 and no suggested replacement. These models can no longer be used as a base model for a fine-tuning job. | 2026-02-25 source, checked 2026-11-12 | Not published |
| meta-llama/Llama-3.2-3B-Instruct-Turbo The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-03-06 source, checked 2026-11-12 | Not published |
| meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: Yes. Models marked "Yes" can be deployed as on-demand dedicated endpoints, either under the listed ID or as the underlying base model of a serving variant (for example, a deprecated -FP8 or -Turbo ID). | 2026-03-06 source, checked 2026-11-12 | Not published |
| mixedbread-ai/Mxbai-Rerank-Large-V2 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-03-06 source, checked 2026-11-12 | Not published |
| moonshotai/Kimi-K2-Instruct-0905 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. The fine-tuning table lists suggested replacement moonshotai/Kimi-K2.6 and removal date 2026-07-29. These models can no longer be used as a base model for a fine-tuning job. | 2026-03-06 source, checked 2026-11-12 | Not published |
| moonshotai/Kimi-K2-Thinking The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. The fine-tuning table lists suggested replacement moonshotai/Kimi-K2.6 and removal date 2026-07-29. These models can no longer be used as a base model for a fine-tuning job. | 2026-03-06 source, checked 2026-11-12 | Not published |
| meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: Yes. Models marked "Yes" can be deployed as on-demand dedicated endpoints, either under the listed ID or as the underlying base model of a serving variant (for example, a deprecated -FP8 or -Turbo ID). | 2026-03-31 source, checked 2026-11-12 | Not published |
| mistralai/Mistral-Small-24B-Instruct-2501 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-04-02 source, checked 2026-11-12 | Not published |
| Qwen/Qwen3-Next-80B-A3B-Instruct The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: Yes. Models marked "Yes" can be deployed as on-demand dedicated endpoints, either under the listed ID or as the underlying base model of a serving variant (for example, a deprecated -FP8 or -Turbo ID). The fine-tuning table lists suggested replacement Qwen/Qwen3.5-122B-A10B and removal date 2026-07-29. These models can no longer be used as a base model for a fine-tuning job. | 2026-04-02 source, checked 2026-11-12 | Not published |
| zai-org/GLM-4.5-Air-FP8 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-04-02 source, checked 2026-11-12 | Not published |
| zai-org/GLM-4.7 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. The fine-tuning table lists suggested replacement zai-org/GLM-5.1 and removal date 2026-07-29. These models can no longer be used as a base model for a fine-tuning job. | 2026-04-02 source, checked 2026-11-12 | Not published |
| ServiceNow-AI/Apriel-1.5-15b-Thinker The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-04-03 source, checked 2026-11-12 | Not published |
| ServiceNow-AI/Apriel-1.6-15b-Thinker The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-04-03 source, checked 2026-11-12 | Not published |
| mistralai/Mixtral-8x7B-Instruct-v0.1 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: Yes. Models marked "Yes" can be deployed as on-demand dedicated endpoints, either under the listed ID or as the underlying base model of a serving variant (for example, a deprecated -FP8 or -Turbo ID). | 2026-04-16 source, checked 2026-11-12 | Not published |
| Qwen/Qwen3-235B-A22B-Thinking-2507 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: Yes. Models marked "Yes" can be deployed as on-demand dedicated endpoints, either under the listed ID or as the underlying base model of a serving variant (for example, a deprecated -FP8 or -Turbo ID). The inference table also lists removal date 2026-03-06 for this id. | 2026-04-16 source, checked 2026-11-12 | Not published |
| Qwen/Qwen3-VL-8B-Instruct The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: Yes. Models marked "Yes" can be deployed as on-demand dedicated endpoints, either under the listed ID or as the underlying base model of a serving variant (for example, a deprecated -FP8 or -Turbo ID). The fine-tuning table lists suggested replacement Qwen/Qwen3.5-9B and removal date 2026-07-29. These models can no longer be used as a base model for a fine-tuning job. | 2026-04-16 source, checked 2026-11-12 | Not published |
| deepseek-ai/DeepSeek-R1 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. The fine-tuning table lists suggested replacement deepseek-ai/DeepSeek-V3.1 and removal date 2026-07-29. These models can no longer be used as a base model for a fine-tuning job. | 2026-05-14 source, checked 2026-11-12 | Not published |
| deepseek-ai/DeepSeek-V3.1 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: Yes. Models marked "Yes" can be deployed as on-demand dedicated endpoints, either under the listed ID or as the underlying base model of a serving variant (for example, a deprecated -FP8 or -Turbo ID). | 2026-05-14 source, checked 2026-11-12 | Not published |
| Qwen/Qwen3-Coder-Next-FP8 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: Yes. Models marked "Yes" can be deployed as on-demand dedicated endpoints, either under the listed ID or as the underlying base model of a serving variant (for example, a deprecated -FP8 or -Turbo ID). | 2026-05-14 source, checked 2026-11-12 | Not published |
| moonshotai/Kimi-K2.5 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. The fine-tuning table lists suggested replacement moonshotai/Kimi-K2.6 and removal date 2026-07-29. These models can no longer be used as a base model for a fine-tuning job. | 2026-05-21 source, checked 2026-11-12 | Not published |
| black-forest-labs/FLUX.1-krea-dev The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-05-27 source, checked 2026-11-12 | Not published |
| Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: Yes. Models marked "Yes" can be deployed as on-demand dedicated endpoints, either under the listed ID or as the underlying base model of a serving variant (for example, a deprecated -FP8 or -Turbo ID). The changelog says to use MiniMaxAI/MiniMax-M2.7 instead. The deprecations table does not name a replacement for this id. | 2026-06-04 source, checked 2026-11-12 | MiniMaxAI/MiniMax-M2.7 (official) |
| mistralai/Voxtral-Mini-3B-2507 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-06-11 source, checked 2026-11-12 | Not published |
| zai-org/GLM-5 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. The fine-tuning table lists suggested replacement zai-org/GLM-5.1 and removal date 2026-07-29. These models can no longer be used as a base model for a fine-tuning job. The changelog says to use zai-org/GLM-5.2 instead. The deprecations table does not name a replacement for this id. | 2026-06-22 source, checked 2026-11-12 | zai-org/GLM-5.2 (official) |
| Qwen/Qwen3.5-397B-A17B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: Yes. Models marked "Yes" can be deployed as on-demand dedicated endpoints, either under the listed ID or as the underlying base model of a serving variant (for example, a deprecated -FP8 or -Turbo ID). The changelog says to use MiniMaxAI/MiniMax-M3 instead. The deprecations table does not name a replacement for this id. | 2026-06-29 source, checked 2026-11-12 | MiniMaxAI/MiniMax-M3 (official) |
| meta-llama/Meta-Llama-3-8B-Instruct-Lite The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-07-10 source, checked 2026-11-12 | Not published |
| Qwen/Qwen3-235B-A22B-Instruct-2507-tput The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: Yes. Models marked "Yes" can be deployed as on-demand dedicated endpoints, either under the listed ID or as the underlying base model of a serving variant (for example, a deprecated -FP8 or -Turbo ID). | 2026-07-10 source, checked 2026-11-12 | Not published |
| zai-org/GLM-5.1 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: Yes. Models marked "Yes" can be deployed as on-demand dedicated endpoints, either under the listed ID or as the underlying base model of a serving variant (for example, a deprecated -FP8 or -Turbo ID). | 2026-07-10 source, checked 2026-11-12 | Not published |
| deepseek-ai/DeepSeek-R1-0528 This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement deepseek-ai/DeepSeek-V3.1 and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | deepseek-ai/DeepSeek-V3.1 (official) |
| deepseek-ai/DeepSeek-R1-Distill-Llama-70B-131k This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement meta-llama/Llama-3.3-70B-Instruct-Reference and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | meta-llama/Llama-3.3-70B-Instruct-Reference (official) |
| deepseek-ai/DeepSeek-R1-Distill-Llama-70B-32k This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement meta-llama/Llama-3.3-70B-Instruct-Reference and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | meta-llama/Llama-3.3-70B-Instruct-Reference (official) |
| deepseek-ai/DeepSeek-V3 This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement deepseek-ai/DeepSeek-V3.1 and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | deepseek-ai/DeepSeek-V3.1 (official) |
| deepseek-ai/DeepSeek-V3-0324 This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement deepseek-ai/DeepSeek-V3.1 and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | deepseek-ai/DeepSeek-V3.1 (official) |
| deepseek-ai/DeepSeek-V3-Base This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement deepseek-ai/DeepSeek-V3.1 and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | deepseek-ai/DeepSeek-V3.1 (official) |
| deepseek-ai/DeepSeek-V3.1-Base This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement deepseek-ai/DeepSeek-V3.1 and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | deepseek-ai/DeepSeek-V3.1 (official) |
| google/gemma-3-12b-it This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement google/gemma-4-26B-A4B-it and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | google/gemma-4-26B-A4B-it (official) |
| google/gemma-3-12b-it-VLM This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement google/gemma-4-31B-it-VLM and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | google/gemma-4-31B-it-VLM (official) |
| google/gemma-3-12b-pt This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement google/gemma-4-26B-A4B-it and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | google/gemma-4-26B-A4B-it (official) |
| google/gemma-3-1b-it This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement google/gemma-4-26B-A4B-it and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | google/gemma-4-26B-A4B-it (official) |
| google/gemma-3-1b-pt This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement google/gemma-4-26B-A4B-it and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | google/gemma-4-26B-A4B-it (official) |
| google/gemma-3-270m This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-0.8B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-0.8B (official) |
| google/gemma-3-270m-it This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-0.8B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-0.8B (official) |
| google/gemma-3-27b-it This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement google/gemma-4-31B-it and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | google/gemma-4-31B-it (official) |
| google/gemma-3-27b-it-VLM This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement google/gemma-4-31B-it-VLM and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | google/gemma-4-31B-it-VLM (official) |
| google/gemma-3-27b-pt This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement google/gemma-4-31B-it and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | google/gemma-4-31B-it (official) |
| google/gemma-3-4b-it This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement google/gemma-4-26B-A4B-it and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | google/gemma-4-26B-A4B-it (official) |
| google/gemma-3-4b-it-VLM This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement google/gemma-4-26B-A4B-it and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | google/gemma-4-26B-A4B-it (official) |
| google/gemma-3-4b-pt This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement google/gemma-4-26B-A4B-it and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | google/gemma-4-26B-A4B-it (official) |
| meta-llama/Llama-3.2-1B This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement meta-llama/Meta-Llama-3.1-8B-Instruct-Reference and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | meta-llama/Meta-Llama-3.1-8B-Instruct-Reference (official) |
| meta-llama/Llama-3.2-1B-Instruct This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement meta-llama/Meta-Llama-3.1-8B-Instruct-Reference and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | meta-llama/Meta-Llama-3.1-8B-Instruct-Reference (official) |
| meta-llama/Llama-3.2-3B This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement meta-llama/Meta-Llama-3.1-8B-Instruct-Reference and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | meta-llama/Meta-Llama-3.1-8B-Instruct-Reference (official) |
| meta-llama/Llama-3.2-3B-Instruct This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement meta-llama/Meta-Llama-3.1-8B-Instruct-Reference and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | meta-llama/Meta-Llama-3.1-8B-Instruct-Reference (official) |
| meta-llama/Llama-3.3-70B-131k-Instruct-Reference This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement meta-llama/Llama-3.3-70B-Instruct-Reference and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | meta-llama/Llama-3.3-70B-Instruct-Reference (official) |
| meta-llama/Llama-3.3-70B-32k-Instruct-Reference This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement meta-llama/Llama-3.3-70B-Instruct-Reference and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | meta-llama/Llama-3.3-70B-Instruct-Reference (official) |
| meta-llama/Llama-4-Maverick-17B-128E This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement meta-llama/Llama-4-Maverick-17B-128E-Instruct and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | meta-llama/Llama-4-Maverick-17B-128E-Instruct (official) |
| meta-llama/Llama-4-Scout-17B-16E This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement meta-llama/Llama-4-Scout-17B-16E-Instruct and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | meta-llama/Llama-4-Scout-17B-16E-Instruct (official) |
| meta-llama/Meta-Llama-3-70B-Instruct This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement meta-llama/Llama-3.3-70B-Instruct-Reference and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | meta-llama/Llama-3.3-70B-Instruct-Reference (official) |
| meta-llama/Meta-Llama-3-8B This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement meta-llama/Meta-Llama-3.1-8B-Instruct-Reference and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | meta-llama/Meta-Llama-3.1-8B-Instruct-Reference (official) |
| meta-llama/Meta-Llama-3.1-405B-10k-Instruct-Reference This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists removal date 2026-07-29 and no suggested replacement. | 2026-07-29 source, checked 2026-11-12 | Not published |
| meta-llama/Meta-Llama-3.1-405B-10k-Reference This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists removal date 2026-07-29 and no suggested replacement. | 2026-07-29 source, checked 2026-11-12 | Not published |
| meta-llama/Meta-Llama-3.1-405B-8k-Instruct-Reference This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists removal date 2026-07-29 and no suggested replacement. | 2026-07-29 source, checked 2026-11-12 | Not published |
| meta-llama/Meta-Llama-3.1-405B-8k-Reference This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists removal date 2026-07-29 and no suggested replacement. | 2026-07-29 source, checked 2026-11-12 | Not published |
| meta-llama/Meta-Llama-3.1-405B-Instruct-Reference This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists removal date 2026-07-29 and no suggested replacement. | 2026-07-29 source, checked 2026-11-12 | Not published |
| meta-llama/Meta-Llama-3.1-405B-Reference This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists removal date 2026-07-29 and no suggested replacement. | 2026-07-29 source, checked 2026-11-12 | Not published |
| meta-llama/Meta-Llama-3.1-70B-131k-Instruct-Reference This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement meta-llama/Llama-3.3-70B-Instruct-Reference and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | meta-llama/Llama-3.3-70B-Instruct-Reference (official) |
| meta-llama/Meta-Llama-3.1-70B-131k-Reference This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement meta-llama/Llama-3.3-70B-Instruct-Reference and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | meta-llama/Llama-3.3-70B-Instruct-Reference (official) |
| meta-llama/Meta-Llama-3.1-70B-32k-Instruct-Reference This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement meta-llama/Llama-3.3-70B-Instruct-Reference and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | meta-llama/Llama-3.3-70B-Instruct-Reference (official) |
| meta-llama/Meta-Llama-3.1-70B-32k-Reference This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement meta-llama/Llama-3.3-70B-Instruct-Reference and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | meta-llama/Llama-3.3-70B-Instruct-Reference (official) |
| meta-llama/Meta-Llama-3.1-70B-Instruct-Reference This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement meta-llama/Llama-3.3-70B-Instruct-Reference and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | meta-llama/Llama-3.3-70B-Instruct-Reference (official) |
| meta-llama/Meta-Llama-3.1-70B-Reference This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement meta-llama/Llama-3.3-70B-Instruct-Reference and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | meta-llama/Llama-3.3-70B-Instruct-Reference (official) |
| meta-llama/Meta-Llama-3.1-8B-131k-Instruct-Reference This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement meta-llama/Meta-Llama-3.1-8B-Instruct-Reference and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | meta-llama/Meta-Llama-3.1-8B-Instruct-Reference (official) |
| meta-llama/Meta-Llama-3.1-8B-131k-Reference This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement meta-llama/Meta-Llama-3.1-8B-Instruct-Reference and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | meta-llama/Meta-Llama-3.1-8B-Instruct-Reference (official) |
| meta-llama/Meta-Llama-3.1-8B-Reference This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement meta-llama/Meta-Llama-3.1-8B-Instruct-Reference and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | meta-llama/Meta-Llama-3.1-8B-Instruct-Reference (official) |
| mistralai/Mistral-7B-Instruct-v0.2 This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement mistralai/Mixtral-8x7B-Instruct-v0.1 and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | mistralai/Mixtral-8x7B-Instruct-v0.1 (official) |
| mistralai/Mixtral-8x7B-v0.1 This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement mistralai/Mixtral-8x7B-Instruct-v0.1 and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | mistralai/Mixtral-8x7B-Instruct-v0.1 (official) |
| moonshotai/Kimi-K2-Base This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement moonshotai/Kimi-K2.6 and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | moonshotai/Kimi-K2.6 (official) |
| moonshotai/Kimi-K2-Instruct This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement moonshotai/Kimi-K2.6 and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | moonshotai/Kimi-K2.6 (official) |
| nvidia/NVIDIA-Nemotron-Nano-9B-v2 This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-9B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-9B (official) |
| Qwen/Qwen2-1.5B This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-2B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-2B (official) |
| Qwen/Qwen2-1.5B-Instruct This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-2B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-2B (official) |
| Qwen/Qwen2-72B This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement meta-llama/Llama-3.3-70B-Instruct-Reference and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | meta-llama/Llama-3.3-70B-Instruct-Reference (official) |
| Qwen/Qwen2-7B This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-9B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-9B (official) |
| Qwen/Qwen2-7B-Instruct This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-9B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-9B (official) |
| Qwen/Qwen2.5-1.5B This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-2B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-2B (official) |
| Qwen/Qwen2.5-1.5B-Instruct This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-2B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-2B (official) |
| Qwen/Qwen2.5-14B-Instruct This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-27B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-27B (official) |
| Qwen/Qwen2.5-32B This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-27B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-27B (official) |
| Qwen/Qwen2.5-32B-Instruct This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-27B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-27B (official) |
| Qwen/Qwen2.5-3B This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-4B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-4B (official) |
| Qwen/Qwen2.5-3B-Instruct This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-4B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-4B (official) |
| Qwen/Qwen2.5-72B This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement meta-llama/Llama-3.3-70B-Instruct-Reference and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | meta-llama/Llama-3.3-70B-Instruct-Reference (official) |
| Qwen/Qwen2.5-72B-Instruct This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement meta-llama/Llama-3.3-70B-Instruct-Reference and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | meta-llama/Llama-3.3-70B-Instruct-Reference (official) |
| Qwen/Qwen2.5-7B-Instruct This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-9B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-9B (official) |
| Qwen/Qwen3-0.6B This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-0.8B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-0.8B (official) |
| Qwen/Qwen3-0.6B-Base This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-0.8B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-0.8B (official) |
| Qwen/Qwen3-1.7B This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-2B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-2B (official) |
| Qwen/Qwen3-1.7B-Base This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-2B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-2B (official) |
| Qwen/Qwen3-14B This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-27B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-27B (official) |
| Qwen/Qwen3-14B-Base This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-27B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-27B (official) |
| Qwen/Qwen3-235B-A22B This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-397B-A17B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-397B-A17B (official) |
| Qwen/Qwen3-235B-A22B-Instruct-2507 This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-397B-A17B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-397B-A17B (official) |
| Qwen/Qwen3-30B-A3B This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.6-35B-A3B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.6-35B-A3B (official) |
| Qwen/Qwen3-30B-A3B-Base This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.6-35B-A3B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.6-35B-A3B (official) |
| Qwen/Qwen3-30B-A3B-Instruct-2507 This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.6-35B-A3B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.6-35B-A3B (official) |
| Qwen/Qwen3-32B This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-27B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-27B (official) |
| Qwen/Qwen3-4B This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-4B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-4B (official) |
| Qwen/Qwen3-4B-Base This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-4B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-4B (official) |
| Qwen/Qwen3-8B This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-9B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-9B (official) |
| Qwen/Qwen3-8B-Base This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-9B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-9B (official) |
| Qwen/Qwen3-Coder-30B-A3B-Instruct This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.6-35B-A3B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.6-35B-A3B (official) |
| Qwen/Qwen3-Coder-480B-A35B-Instruct This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists removal date 2026-07-29 and no suggested replacement. | 2026-07-29 source, checked 2026-11-12 | Not published |
| Qwen/Qwen3-VL-235B-A22B-Instruct This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists removal date 2026-07-29 and no suggested replacement. | 2026-07-29 source, checked 2026-11-12 | Not published |
| Qwen/Qwen3-VL-30B-A3B-Instruct This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-4B and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | Qwen/Qwen3.5-4B (official) |
| togethercomputer/llama-2-7b-chat This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement meta-llama/Meta-Llama-3.1-8B-Instruct-Reference and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | meta-llama/Meta-Llama-3.1-8B-Instruct-Reference (official) |
| zai-org/GLM-4.6 This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement zai-org/GLM-5.1 and removal date 2026-07-29. | 2026-07-29 source, checked 2026-11-12 | zai-org/GLM-5.1 (official) |
| black-forest-labs/FLUX.1-schnell The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-08-19 source, checked 2026-11-12 | Not published |
| moonshotai/Kimi-K2.6 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: Yes. Models marked "Yes" can be deployed as on-demand dedicated endpoints, either under the listed ID or as the underlying base model of a serving variant (for example, a deprecated -FP8 or -Turbo ID). | 2026-08-19 source, checked 2026-11-12 | Not published |
| Qwen/Qwen2.5-7B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-9B and removal date 2026-07-29. These models can no longer be used as a base model for a fine-tuning job. | 2026-08-19 source, checked 2026-11-12 | Not published |
| deepcogito/cogito-v2-1-671b The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-08-21 source, checked 2026-11-12 | Not published |
| google/gemma-3n-E4B-it The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-08-25 source, checked 2026-11-12 | Not published |
| meta-llama/Llama-Guard-4-12B The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-08-25 source, checked 2026-11-12 | Not published |
| deepseek-ai/DeepSeek-V4-Pro The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: Yes. Models marked "Yes" can be deployed as on-demand dedicated endpoints, either under the listed ID or as the underlying base model of a serving variant (for example, a deprecated -FP8 or -Turbo ID). The changelog says to use deepseek-ai/DeepSeek-V4-Pro-0813 instead. The deprecations table does not name a replacement for this id. | 2026-08-27 source, checked 2026-11-12 | deepseek-ai/DeepSeek-V4-Pro-0813 (official) |
| moonshotai/Kimi-K2.7-Code The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: Yes. Models marked "Yes" can be deployed as on-demand dedicated endpoints, either under the listed ID or as the underlying base model of a serving variant (for example, a deprecated -FP8 or -Turbo ID). | 2026-08-27 source, checked 2026-11-12 | Not published |
| nvidia/Nemotron-3-ultra-550b-a55b The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-08-27 source, checked 2026-11-12 | Not published |
| pearl-ai/gemma-4-31b-it The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-08-27 source, checked 2026-11-12 | Not published |
| nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16 This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists removal date 2026-09-11 and no suggested replacement. | 2026-09-11 source, checked 2026-11-12 | Not published |
| nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16 This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists removal date 2026-09-11 and no suggested replacement. | 2026-09-11 source, checked 2026-11-12 | Not published |
| Qwen/Qwen3.5-35B-A3B-Base This row is the fine-tuning service, not serverless inference. The inference table does not list this id. These models can no longer be used as a base model for a fine-tuning job. The fine-tuning table lists suggested replacement Qwen/Qwen3.5-35B-A3B and removal date 2026-09-11. | 2026-09-11 source, checked 2026-11-12 | Qwen/Qwen3.5-35B-A3B (official) |
| google/gemma-4-31B-it The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: Yes. Models marked "Yes" can be deployed as on-demand dedicated endpoints, either under the listed ID or as the underlying base model of a serving variant (for example, a deprecated -FP8 or -Turbo ID). Sources disagree. The scheduled deprecations table lists removal date 2026-09-14. The inference table lists removal date 2026-09-15. The stored shutdown date is the inference table date. | 2026-09-15 source, checked 2026-11-12 | zai-org/GLM-5.3-Flash (official) |
| intfloat/multilingual-e5-large-instruct The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. Sources disagree. The scheduled deprecations table lists removal date 2026-09-14. The inference table lists removal date 2026-09-15. The stored shutdown date is the inference table date. | 2026-09-15 source, checked 2026-11-12 | Not published |
| openai/gpt-oss-20b The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: Yes. Models marked "Yes" can be deployed as on-demand dedicated endpoints, either under the listed ID or as the underlying base model of a serving variant (for example, a deprecated -FP8 or -Turbo ID). Sources disagree. The scheduled deprecations table lists removal date 2026-09-14. The inference table lists removal date 2026-09-15. The stored shutdown date is the inference table date. | 2026-09-15 source, checked 2026-11-12 | Qwen/Qwen3.5-9B (official) |
| thinkingmachines/Inkling-Small The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: Yes. Models marked "Yes" can be deployed as on-demand dedicated endpoints, either under the listed ID or as the underlying base model of a serving variant (for example, a deprecated -FP8 or -Turbo ID). Sources disagree. The scheduled deprecations table lists removal date 2026-09-14. The inference table lists removal date 2026-09-15. The stored shutdown date is the inference table date. | 2026-09-15 source, checked 2026-11-12 | zai-org/GLM-5.3-Flash (official) |
| deepseek-ai/DeepSeek-V4-Flash-0731 The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: Yes. Models marked "Yes" can be deployed as on-demand dedicated endpoints, either under the listed ID or as the underlying base model of a serving variant (for example, a deprecated -FP8 or -Turbo ID). The changelog says to use deepseek-ai/DeepSeek-V4.1-Flash-1017 instead. The deprecations table does not name a replacement for this id. | 2026-10-22 source, checked 2026-11-12 | deepseek-ai/DeepSeek-V4.1-Flash-1017 (official) |
| meta-llama/Llama-3.3-70B-Instruct-Turbo The inference table lists this id as removed from serverless inference. Supported by on-demand dedicated endpoints: Yes. Models marked "Yes" can be deployed as on-demand dedicated endpoints, either under the listed ID or as the underlying base model of a serving variant (for example, a deprecated -FP8 or -Turbo ID). The changelog says to use meta-models/Muse-Glimmer-30B instead. The deprecations table does not name a replacement for this id. The changelog says for Llama 3.3 70B, deploy meta-llama/Llama-3.3-70B-Instruct on a dedicated endpoint. | 2026-10-22 source, checked 2026-11-12 | meta-models/Muse-Glimmer-30B (official) |
| Qwen/Qwen3.8-2.4T-A95B The scheduled deprecations table says this model will be removed from serverless inference on the date listed. Supported by on-demand dedicated endpoints: No. Models marked "No" are not available as on-demand endpoints and require migration to a different model or a monthly reserved dedicated endpoint. | 2026-11-04 source, checked 2026-11-12 | Qwen/Qwen3.7-Max (official) |
| MiniMaxAI/MiniMax-M3 The scheduled deprecations table says this model will be removed from serverless inference on the date listed. Supported by on-demand dedicated endpoints: Yes. Models marked "Yes" can be deployed as on-demand dedicated endpoints, either under the listed ID or as the underlying base model of a serving variant (for example, a deprecated -FP8 or -Turbo ID). | 2026-11-06 source, checked 2026-11-12 | MiniMaxAI/MiniMax-M3.1-Flash (official) |
| thinkingmachines/Inkling The scheduled deprecations table says this model will be removed from serverless inference on the date listed. Supported by on-demand dedicated endpoints: Yes. Models marked "Yes" can be deployed as on-demand dedicated endpoints, either under the listed ID or as the underlying base model of a serving variant (for example, a deprecated -FP8 or -Turbo ID). | 2026-11-06 source, checked 2026-11-12 | meta-models/Muse-Spark-Open-816B-A42B (official) |
315 Together AI model records have status retired.
The latest shutdown date on a retired model record is 2026-11-06. 2 retired model records share that date, including MiniMaxAI/MiniMax-M3 (Together AI).
92 of 315 retired model records list a replacement. The table links each replacement page.
The official page is https://docs.together.ai/docs/deprecations. Each model page records the day that page was checked. Dataset version 2026-11-10.