MiniMax
MiniMax-M2 redefines efficiency for agents. It is a compact, fast, and cost-effective MoE model (230 billion total parameters with 10 billion active parameters) built for elite performance in coding and agentic tasks, all while maintaining powerful general intelligence.
MiniMax model releasesRank #100 across 526
Rank #140 across 436
Rank #141 across 357
Percentile score by analysis domain.
* Cost is inverted: lower input, output, and blended prices rank higher.
Higher bars mean stronger relative placement.
| Metric | Domain | Value | Rank |
|---|---|---|---|
| Artificial Analysis Intelligence Index | overall | 36.1 | #100 |
| Artificial Analysis Coding Index | coding | 29.2 | #140 |
| Artificial Analysis Math Index | math | 78.3 | #74 |
| MMLU-Pro | reasoning | 82.0% | #77 |
| reasoning |
| 77.7% |
| #140 |
| Humanity's Last Exam | reasoning | 12.5% | #133 |
| LiveCodeBench | coding | 82.6% | #21 |
| SciCode | coding, reasoning | 36.1% | #188 |
| Output Speed | speed | 102.9 tok/s | #132 |
| Time to First Token | speed | 1.06s | #135 |
| Blended Price | cost | $0.525/M | #141 |
| Input Price | cost | $0.300/M | #150 |
| Output Price | cost | $1.20/M | #144 |
| Value Index | cost, overall | 68.8 | #95 |