MiniMax-M2.5 — benchmark results
MiniMax's open M2.5 MoE (230B total, 10B active) for cost-efficient coding and agentic work. Provider: MiniMax. Released 2026-01-15. Access: Open.
Unified ELO 1649 ± 11, rank #368 of 1776 rated models, from 264 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| ActTraitBench | 2.17 | G_KD (Global Knowledge-Decision Gap) (self-reported) | 100 |
| AA TAU-2 Bench | 95.32 | Accuracy (%) | 94.5 |
| GSMA Open-Telco - srsRAN-Bench | 84.67 | Score (%) | 93.6 |
| OpenEvals - SWE-bench Verified | 75.8 | Resolved (%) | 92.9 |
| GSM-MC | 99.39 | Accuracy (%) | 91.8 |
| PinTrace | 45.42 | Ï_U(%) (self-reported) | 88.9 |
| AA IFBench | 71.63 | Accuracy (%) | 88.3 |
| AA-LCR | 66 | Score (self-reported) | 87 |
| UGI - Natural Intelligence | 38.64 | NatInt Score | 86.6 |
| MATH-MC Level 1 | 98.84 | Accuracy (%) | 85.3 |
| JudgeBench Math | 92.86 | Accuracy (%) | 84.3 |
| AA Long Context Reasoning | 66 | Accuracy (%) | 84.1 |
Interactive version: theaggregate.ai/model?slug=minimax-m2-5 · How the rankings work · Data refreshed daily, snapshot 2026-07-22.