MiMo-V2-Pro: benchmark results
Xiaomi's flagship MiMo-V2 Pro MoE model, served API-only. Provider: Xiaomi. Released 2026-03-18. Access: API.
Aggregate Rating 1715 ± 10, rank #168 of 1656 rated models, from 246 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Altered Riddles | 0.29 | Conditioned Override Rate | 100 |
| DORA - Temporal Evolution Reasoning | 63.58 | Final-answer accuracy (%; ReAct agent with tools) | 100 |
| Position Bias (Lechmazur) | 19.8 | Order Flip % (lower is better) | 100 |
| BenchTable - Tech | 89.9 | Weighted Score (%) | 98 |
| Tau2-Bench Telecom | 95.03 | Success Rate (%) | 93.9 |
| AA TAU-2 Bench | 95.03 | Accuracy (%) | 93.8 |
| BenchTable - Reasoning | 77.3 | Weighted Score (%) | 92.9 |
| AA-Omniscience Hallucination Rate | 29.98 | Hallucination Rate (%) | 92.6 |
| IKP - T2 | 100 | Accuracy (%) | 92.6 |
| DORA - Disaster Operational Planning | 55.48 | Final-answer accuracy (%; ReAct agent with tools) | 91.7 |
| DSAEval | 7.91 | Total score (0-10; mean of two LLM judges) | 91.7 |
| BenchTable - STEM | 77.3 | Weighted Score (%) | 88.7 |
Interactive version: theaggregate.ai/model?slug=mimo-v2-pro · How It Works · Data refreshed daily, snapshot 2026-10-20.