See key specs and per-benchmark scores for each model/mode. Scroll horizontally for all columns. 当前对比 2 个模型的评测数据与核心参数。

MiniMax M2
MiniMaxAI
Each axis is a category average, normalized to a 100-point radar.
Relative edge: 指令跟随 +2.3 / Relative gap: AI Agent - 信息收集 -32.3
Relative edge: AI Agent - 信息收集 +32.3 / Relative gap: 指令跟随 -2.3
Method: for each model and benchmark, the chart first averages all scores in the current mode scope instead of taking the best score, then averages those benchmark scores within each category. Only benchmarks with at least two selected models scored are included; missing values are not counted as zero.
Best overall
MiniMax M2.5 · 73.60
Best single
MiniMax M2.5 · τ²-Bench - Telecom 97.80
Modality coverage
MiniMax M2 · 1 modalities
Head to head
7
Benchmarks
1
Wins
6
Losses
-10.57
Average diff
Compare benchmark results across thinking modes and tool usage.
Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology
Complete scores for each model/mode across selected benchmarks.
7 benchmarks with comparable scores. Each model shows its best score; mode label is displayed below.
| Benchmark | MiniMax M2 | MiniMax M2.5 |
|---|---|---|
GPQA Diamond 综合评估 | 78.00Thinking Enabled | 85.20Thinking Enabled |
HLE 综合评估 | 12.50Thinking Enabled | 19.40Thinking Enabled |
SWE-bench Verified 编程与软件工程 | 69.40Thinking Enabled | Tools | 80.20Thinking Enabled | Tools |
AIME2025 数学推理 | 78.00Thinking Enabled | 86.30Thinking Enabled |
τ²-Bench - Telecom Agent能力评测 | 87.00Thinking Enabled | Tools | 97.80Thinking Enabled | Tools |
IF Bench 指令跟随 | 72.30Thinking Enabled | 70.00Thinking Enabled | Tools |
BrowseComp AI Agent - 信息收集 | 44.00Thinking Enabled | Tools | 76.30Thinking Enabled | Tools |
Side-by-side input/output token pricing
Licensing, MoE architecture, and multi-modality support.
| Features & specs | MiniMax M2MiniMaxAI | MiniMax M2.5MiniMaxAI |
|---|---|---|
Core specsRelease | 2025-10-27 | 2026-02-12 |
Context length | 205K | 128K |
Parameters | 2300 | 2290 |
Active parameters | 100 | 100 |
MoE | Yes | Yes |
Supported modes | 常规模式(Non-Thinking Mode)思考模式(Thinking Mode) | No mode data |
LicenseCode Open Source | Closed Source | Closed Source |
Weights Open Source | Closed Source | Not provided |
Commercial use | 免费商用授权 | 免费商用授权 |
Modality supportText Input/Output | / | / |
ResourcesPaper / report | MiniMax M2.5: Built for Real-World Productivity. | |
DataLearner blog | MiniMaxAI开源MiniMax M2模型:Artificial Analysis评测显示综合智能得分超过Claude Opus 4.1,开源第一,全球第五。 | Not provided |

MiniMax M2.5
MiniMaxAI