MiniMax-M3 (Reasoning): benchmark results
Provider: MiniMax. Released 2026-06-01. Access: Open.
Aggregate Rating 1600 ± 16, rank #766 of 2287 rated models, from 140 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| SEA-HELM (Burmese) - Sentiment Analysis | 48.67 | Normalized Score | 95.9 |
| SEA-HELM (Indonesian) - Safety | 64.1 | Normalized Score | 94.6 |
| SEA-HELM (Indonesian) - Toxicity Detection | 56 | Normalized Score | 94.6 |
| SEA-HELM (Tamil) - SEA-Safeguard | 84.64 | Normalized Score | 94.6 |
| SEA-HELM (Tamil) - Natural Language Inference | 55.35 | Normalized Score | 93.2 |
| SEA-HELM (Burmese) - NLU | 64.61 | Normalized Score | 91.9 |
| SEA-HELM (Tamil) - NLR | 74.78 | Normalized Score | 89.2 |
| SEA-HELM (Vietnamese) - SEA-Safeguard | 80.31 | Normalized Score | 89.2 |
| SEA-HELM (Indonesian) - Pragmatics | 74.72 | Normalized Score | 87.8 |
| SEA-HELM (Indonesian) - Sentiment Analysis | 80.73 | Normalized Score | 87.8 |
| SEA-HELM (Tamil) - SEA-MT-Bench (LLM Judge) | 82.19 | Normalized Score | 87.8 |
| SEA-HELM (Tamil) | 72.89 | Normalized Score | 86.5 |
Interactive version: theaggregate.ai/model?slug=minimax-m3-reasoning · How It Works · Data refreshed daily, snapshot 2026-11-22.