UTC

MiniMax-M3

MiniMax · open weights · served by 66 providers · tool calling, reasoning, image input, video input

minimax/MiniMax-M3

Weights
Open
Released
2026-06-01
Knowledge cutoff
2025-01
Context window
1M
Deprecation dates

When each platform deprecates MiniMax-M3, as LiteLLM records it. Plan the move before the date: after it, the platform may stop serving the model or stop updating it.

  • Together AIdeprecated 2026-11-06together_ai/MiniMaxAI/MiniMax-M3successor: together_ai/MiniMaxAI/MiniMax-M3.1-Flash
Every upcoming deprecation →
Hugging Face repository

MiniMaxAI/MiniMax-M3 at commit d95e30cb91, last changed 2026-10-26

Licence
minimax-community
Access
Open download
Library
transformers
Formats
.safetensors
  • Loading it runs code from the repository (trust_remote_code). Read that code, and pin the commit.
  • No model signature file is published with the weights.

What the Hub's scanners report

33 weight files, 33 with a scan result. No scanner reports an issue with them.

The model at this commit, its 33 weight files with their SHA-256.

Facts from the Hugging Face Hub and its file scanners (Protect AI, Cisco, Hugging Face picklescan, VirusTotal, JFrog), as of 2026-11-18. ShieldMap does not copy model cards.

Who serves it
  • 302.AIMiniMax-M3
  • AbacusMiniMaxAI/MiniMax-M3
  • AbliteratedLLMminimax-m3
  • Adzuki AIminimax-m3
  • AIHubMixminimax-m3
  • Ambientminimax/minimax-m3
  • Boundlessminimax-m3
  • ClinePasscline-pass/minimax-m3
  • CommandCodeMiniMaxAI/MiniMax-M3
  • Cortecsminimax-m3
  • CrossModelminimax/minimax-m3
  • Deep InfraMiniMaxAI/MiniMax-M3
  • Eden AIminimax/MiniMax-M3
  • EmpirioLabs AIminimax-m3
  • Fireworks AIaccounts/fireworks/models/minimax-m3
  • Fireworks AIaccounts/fireworks/routers/minimax-latest
  • FrogBotminimax-m3
  • GMI CloudMiniMaxAI/MiniMax-M3
  • Hugging FaceMiniMaxAI/MiniMax-M3
  • HUMAINhumain-m3-previewbeta
  • HUMAINhumain-m3-research-previewbeta
  • Charm Hyperminimax-m3
  • Incominimax-m3
  • Incominimax-m3:fast
42 more providers

Inference APIs (MiniMaxAI/MiniMax-M3), IteraCompute (minimax/minimax-m3), Jalapeno Cloud (MiniMax-M3), Kenari (minimax-m3), Kilo Gateway (minimax/minimax-m3), DevPass (LLM Gateway) (minimax-m3), LLM Gateway (deepinfra/minimax-m3), LLM Gateway (minimax/minimax-m3), LLM Gateway (novita/minimax-m3), LLM Gateway (tencent/minimax-m3), Merge Gateway (minimax/minimax-m3), Modular (minimax/minimax-m3), NanoGPT (minimax/minimax-m3), NanoGPT (minimax/minimax-m3:thinking), Nebius Token Factory (MiniMaxAI/MiniMax-M3), Novita AI (minimax/minimax-m3), Ofox (minimax/minimax-m3), Ollama Cloud (minimax-m3), OpenCode (minimax-m3), OpenCode (minimax-m3-free), OpenCode Go (minimax-m3), OpenRouter (minimax/minimax-m3), Opper (minimax-m3), OrcaRouter (minimax/minimax-m3), Requesty (minimax-m3), Requesty (minimax-m3@eu), SCNet Token Plan (MiniMax-M3), SiliconFlow (MiniMaxAI/MiniMax-M3), Standard Compute (minimax/minimax-m3), SteadyRouter (MiniMax-M3), Tencent Token Plan (minimax-m-3-0), Tencent Token Plan (minimax-m3), TensorX (minimax/minimax-m3), Together AI (MiniMaxAI/MiniMax-M3), Vancine (MiniMax-M3), Vercel AI Gateway (minimax/minimax-m3), Volcengine Ark Coding Plan (minimax-m3), Vultr (minimax-m3), Wally (minimax-m3), CoreWeave (MiniMaxAI/MiniMax-M3), Watt Cloud (MiniMax-M3), ZenMux (minimax/minimax-m3)

Running it yourself

Open weights are loaded and served by code that has its own advisories. These inference servers and model loaders are on ShieldMap; load weights only from the publisher's own repository, and prefer safetensors over formats that can run code.

How supply-chain attacks on models work →

Sources: models.dev (MIT) and the LiteLLM model catalogue (MIT). Deprecation dates are as each platform published them and as LiteLLM records them; check the platform before you rely on one. Data as of 2026-11-22.