minimax-m2-her: benchmark results
MiniMax's MiniMax-M2, a 230B-parameter (10B active) open-weight mixture-of-experts model built for coding and agentic workflows. Provider: MiniMax. Released 2025-10-23. Access: API.
Aggregate Rating 1337 ± 16, rank #1723 of 2170 rated models, from 88 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AGC-Bench - science_analogies | 2.23 | Dataset z-score | 97.5 |
| AGC-Bench - unfun_corpus | 0.75 | Dataset z-score | 94.9 |
| AGC-Bench - story_generation_rocstories | 1.21 | Dataset z-score | 90.2 |
| AGC-Bench - futuregen | 0.87 | Dataset z-score | 83.8 |
| AGC-Bench - permpst | 0.55 | Dataset z-score | 76.2 |
| AGC-Bench - creatset | 1.02 | Dataset z-score | 74.4 |
| AGC-Bench - ocw | 0.58 | Dataset z-score | 67.1 |
| AGC-Bench - grapheval_iclr | 0.47 | Dataset z-score | 65.9 |
| AGC-Bench - grapheval_ai_researcher | 0.26 | Dataset z-score | 58 |
| AGC-Bench - historical_analogy | -0.01 | Dataset z-score | 56.2 |
| AGC-Bench - ttcw | -0.02 | Dataset z-score | 50.6 |
| AGC-Bench - metaphoric_analogies | -0.46 | Dataset z-score | 41.5 |
Interactive version: theaggregate.ai/model?slug=minimax-m2-her · How It Works · Data refreshed daily, snapshot 2026-11-05.