UTC

MiniMax-M2.5

MiniMax · open weights · served by 41 providers · tool calling, reasoning

minimax/MiniMax-M2.5

Weights
Open
Released
2026-02-12
Knowledge cutoff
2025-01
Context window
205K
Deprecation dates

When each platform deprecates MiniMax-M2.5, as LiteLLM records it. Plan the move before the date: after it, the platform may stop serving the model or stop updating it.

  • Basetendeprecated 2026-10-16baseten/MiniMaxAI/MiniMax-M2.5
Every upcoming deprecation →
Hugging Face repository

MiniMaxAI/MiniMax-M2.5 at commit 2515a3ea0a, last changed 2026-10-26

Licence
modified-mit
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

14 weight files, 14 with a scan result. No scanner reports an issue with them.

The model at this commit, its 14 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-M2.5
  • Alibaba Token PlanMiniMax-M2.5
  • Alibaba Token Plan (China)MiniMax-M2.5
  • Amazon Bedrockminimax.minimax-m2.5
  • Clarifaiminimaxai/chat-completion/models/MiniMax-M2_5-high-throughput
  • CloudFerro SherlockMiniMaxAI/MiniMax-M2.5
  • CommandCodeMiniMaxAI/MiniMax-M2.5
  • Cortecsminimax-m2.5
  • Deep InfraMiniMaxAI/MiniMax-M2.5deprecated
  • DInferenceminimax-m2.5deprecated
  • D.Run (China)public/minimax-m25
  • Eden AIminimax/MiniMax-M2.5
  • FrogBotminimax-m2-5deprecated
  • GreenPTminimax-m2.5
  • HPC-AIminimax/minimax-m2.5
  • Hugging FaceMiniMaxAI/MiniMax-M2.5
  • IteraComputeminimax/minimax-m2.5
  • Kilo Gatewayminimax/minimax-m2.5
  • DevPass (LLM Gateway)minimax-m2.5
  • LLM Gatewayminimax/minimax-m2.5
  • LLM Gatewaynovita/minimax-m2.5
  • MeganovaMiniMaxAI/MiniMax-M2.5
  • Merge Gatewayminimax/minimax-m2.5
  • NanoGPTminimax/minimax-m2.5
17 more providers

Novita AI (minimax/minimax-m2.5), Ofox (minimax/minimax-m2.5), OpenCode (minimax-m2.5), OpenCode (minimax-m2.5-free), OpenRouter (minimax/minimax-m2.5), OrcaRouter (minimax/minimax-m2.5), Qiniu (minimax/minimax-m2.5), SCNet Token Plan (MiniMax-M2.5), SiliconFlow (MiniMaxAI/MiniMax-M2.5), stepping stone AG (MiniMaxAI/MiniMax-M2.5), Tencent Coding Plan (China) (minimax-m2.5), TensorX (minimax/minimax-m2.5), Together AI (MiniMaxAI/MiniMax-M2.5), TokenGo (minimax/minimax-m2.5), Venice AI (minimax-m25), Vercel AI Gateway (minimax/minimax-m2.5), ZenMux (minimax/minimax-m2.5)

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.