MiniMax-M2
MiniMax-M2 is an open-weights language model from MiniMax, released October 27, 2025. It has 229B parameters (mixture-of-experts, 11B active). The context window is 200K tokens. API price: $0.3 input / $1.2 output per million tokens.
Today, we are officially open-sourcing and launching MiniMax M2, a model born for Agents and code. At only 8% of the price of Claude Sonnet and twice the speed, it's available for free for a limited time! Top-tier Coding Capabilities: Built for end-to-end development workflows, it excels in various applications such …
Check it against your own PC →
Can you run MiniMax-M2 locally?
The smallest common GPU it runs on is 32 GB of VRAM (with 32 GB of RAM): IQ1_M, GPU + experts in RAM, roughly 34 tokens per second. Estimates assume 32 GB of DDR5 RAM, a 16K context and llama.cpp-style GGUF quantization. Enter your exact hardware in the live app.
| GPU memory | Verdict | Best quant | VRAM | RAM | tok/s |
|---|---|---|---|---|---|
| 6 GB VRAM | Too large | – | – | – | – |
| 8 GB VRAM | Too large | – | – | – | – |
| 12 GB VRAM | Too large | – | – | – | – |
| 16 GB VRAM | Too large | – | – | – | – |
| 24 GB VRAM | Too large | – | – | – | – |
| 32 GB VRAM | Runs · GPU + CPU experts | IQ1_M (experimental) | 31.2 GB | 23.6 GB | ~34 |
Ready-made quantizations on Hugging Face: unsloth/MiniMax-M2-GGUF · bartowski/MiniMaxAI_MiniMax-M2-GGUF · redponike/MiniMax-M2-GGUF · lmstudio-community/MiniMax-M2-GGUF · DevQuasar/MiniMaxAI.MiniMax-M2-GGUF · cturan/MiniMax-M2-GGUF
Benchmark results
Epoch AI has not published results for MiniMax-M2 yet. New models are usually evaluated within days to weeks.
Links
Announcement · Hugging Face: MiniMaxAI/MiniMax-M2
More from MiniMax
MiniMax-M3.1-Flash · MiniMax-M3 · MiniMax-M2.7 · MiniMax-M2.5 · MiniMax-M2.1 · MiniMax-M1-80k · MiniMax-M1-40k