M2

M2.1

Chat modelMiniMax M2.1

MiniMax M2.1 Preview

Release date: 2025-12-23Updated: 2026-03-08 21:11:182,690
Parameters
230B
Context length
200K
Chinese support
Supported
Reasoning ability

MiniMax M2.1 Preview is an AI model published by MiniMaxAI, released on 2025-12-23, for Chat model, with 230B parameters, and 200K context length, under the Modified MIT license, with a 88.00 score on MMLU Pro.

Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology

M2.1

Model basics

Reasoning traces
Supported
Thinking modes
Standard Mode
Context length
200K tokens
Max output length
128K tokens
Model type
Chat model
Modality (in / out)
Text → Text
Release date
2025-12-23
Model file size
No data
MoE architecture
Yes
Total params / Active params
230B / 10B
Knowledge cutoff
No data
M2.1

Open source & experience

M2.1

Official resources

M2.1

API details

API speed
3/5
No public API pricing yet.
M2.1

Benchmark Results

M2.1 currently shows benchmark results led by MMLU Pro (7 / 126, score 88), SWE-bench Verified (35 / 108, score 74.80), GPQA Diamond (70 / 179, score 81). This page also consolidates core specs, context limits, and API pricing so you can evaluate the model from benchmark results and deployment constraints together.

Thinking
Tool usage

General Knowledge

3 evaluations
Benchmark / mode
Score
Rank/total
88
7 / 126
81
70 / 179
22
96 / 159

Coding and Software Engineer

2 evaluations
Benchmark / mode
Score
Rank/total

Math and Reasoning

1 evaluations
Benchmark / mode
Score
Rank/total
81
56 / 106

Agent Level Benchmark

2 evaluations
Benchmark / mode
Score
Rank/total

Instruction Following

1 evaluations
Benchmark / mode
Score
Rank/total
70
12 / 29

AI Agent - Information Search

1 evaluations
Benchmark / mode
Score
Rank/total
47.40
37 / 45

AI Agent - Tool Usage

1 evaluations
Benchmark / mode
Score
Rank/total
47.90
35 / 46

Claw-style Agent Evaluation

1 evaluations
Benchmark / mode
Score
Rank/total
Pinch Bench
Thinking ModeTools
84.30
18 / 37

Compare with other models

No curated comparisons for this model yet.

Want a custom combination? Open the compare tool

M2.1

Publisher

MiniMax M2.1 Preview

Model Overview

MiniMax M2.1 Preview is an AI model published by MiniMaxAI, released on 2025-12-23, for Chat model, with 230B parameters, and 200K context length, under the Modified MIT license, with a 88.00 score on MMLU Pro.

DataLearner on WeChat

Follow DataLearner on WeChat for AI model updates and research notes.

DataLearner WeChat QR code