MiMo-V2.6-Pro
Grok 4.7
- Context
- 205Ktokens
- Released
- 1y ago
- Size
- 229Bparameters
Efficient open MiniMax model built for coding agents and tool-heavy workflows
- Language model
- Open weights · 229B
- Reasoning
- Tools
How it compares
Level
FlagshipThe lab's most capable generally available line, priced for hard tasks.
Flagship models from other labs
Vs. your stack
Add the models you use to see how this one compares.
Sign inAdditional source evidence
- Source model
- minimax/MiniMax-M2
- Source family
- minimax
- Last updated
- 2025-10-27
- Model context
- 205K
- Model output limit
- 131K
- Open weights
- Yes
- File attachments
- No
- Reasoning
- Yes
- Tool calling
- Yes
- Temperature control
- Yes
Downloadable weights
Published benchmarks
| Benchmark | Score | Evaluation details | Evidence |
|---|---|---|---|
| SWE-Bench Verified | 69.4 · resolved | — | Source |
Complete source record (JSON)
{
"id": "minimax/MiniMax-M2",
"name": "MiniMax-M2",
"description": "Efficient open MiniMax model built for coding agents and tool-heavy workflows",
"family": "minimax",
"last_updated": "2025-10-27",
"attachment": false,
"reasoning": true,
"tool_call": true,
"temperature": true,
"modalities": {
"input": [
"text"
],
"output": [
"text"
]
},
"limit": {
"context": 204800,
"output": 131072
},
"benchmarks": [
{
"name": "SWE-Bench Verified",
"score": 69.4,
"source": "https://huggingface.co/MiniMaxAI/MiniMax-M2",
"metric": "resolved"
}
],
"release_date": "2025-10-27",
"open_weights": true,
"weights": [
{
"label": "Hugging Face",
"url": "https://huggingface.co/MiniMaxAI/MiniMax-M2"
}
]
}Frequently asked questions
Running locally
An estimated 159.9 GB at Q4, 279.9 GB at Q8 and 526.5 GB at FP16. These figures are estimates from the 229B parameter count, covering weights plus about 15% and 0.5 GB for a short context.
No. It does not fit on a 24 GB, 16 GB, 12 GB or 8 GB GPU at Q4, Q8 or FP16, according to estimates.
Yes, on a 256 GB Mac. Estimated VRAM at Q4 is 159.9 GB, and assuming macOS lets the GPU use about 75% of unified memory, 256 GB is the smallest common Mac memory size that fits (estimate).
Privacy and data
Yes. MiniMax-M2 has open weights, so it can be self-hosted.
Capabilities and specs
The context window is 205K tokens (204,800) and the maximum output is 131K tokens (131,072).
No. Its input modalities are text only.
Yes. Tool calling (function calling) is supported.
Yes. A reasoning mode is supported.
Quality and benchmarks
Rank 136 on Arena WebDev with a score of 1299, and rank 232 on Arena Text with a score of 1344.
olmOCR 2 by Ai2, PaddleOCR-VL by Baidu and ERNIE 4.5 VL 28B A3B Thinking by Baidu. They are matched on model kind and release date, with no comparable price.
olmOCR 2MiniMax-M2 vs olmOCR 2PaddleOCR-VLMiniMax-M2 vs PaddleOCR-VLERNIE 4.5 VL 28B A3B ThinkingMiniMax-M2 vs ERNIE 4.5 VL 28B A3B Thinking
Versions and lineage
MiniMax-M2 has a 205K token (204,800) context window and was released October 27, 2025, while MiniMax-M2 Her has a 66K token (65,536) context window and was released January 23, 2026.
Licensing and access
MiniMax-M2 has open weights.
From the Hugging Face repository MiniMaxAI/MiniMax-M2.
Basics
MiniMax made MiniMax-M2, and it was released on October 27, 2025.
MiniMax-M2 has open weights.
Generated from Vertical Tabs data, updated November 11, 2026.
Discussion
No comments yet
Be the first to share your thoughts.