MiMo-V2.5 (Non-reasoning): benchmark results
Provider: Xiaomi. Released 2026-04-23. Access: Open.
Aggregate Rating 1621 ± 29, rank #641 of 2170 rated models, from 19 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| UAV-DualCog - Landmark Visibility Counting | 58.5 | Visibility-count accuracy (%; 1,024 flight videos in which t | 92.9 |
| UAV-DualCog - Flight Behavior Recognition (Composite) | 35.9 | Composite-level behavior accuracy (%; 1,024 first-person UAV | 90.4 |
| LLM Milgram Obedience - Baseline Full-Obedience Rate | 0 | Valid baseline sessions ending at 450 V (%; lower is less ob | 86.8 |
| UGI - Writing | 44.76 | Writing Score | 83.4 |
| UGI - Natural Intelligence | 32.08 | NatInt Score | 76.2 |
| PatternEval - Response-Pattern Failure Rate (Non-Thinking) | 24.92 | Bad-pattern trigger rate (%, lower is better; share of final | 75 |
| UAV-DualCog - Landmark-Relative Direction | 51.2 | Answer accuracy (%; 1,024 questions asking where a landmark | 74.3 |
| UAV-DualCog - Self-Relative Position | 33.1 | Answer accuracy (%; 1,024 questions asking where the UAV is | 74.3 |
| UAV-DualCog - Landmark-Driven Action Decision | 49.7 | Answer accuracy (%; 1,024 questions asking which way the UAV | 72.9 |
| UAV-DualCog - Flight Behavior Recognition (Atomic) | 30.2 | Atomic-level behavior accuracy (%; the atomic flight actions | 69.2 |
| LLM Milgram Obedience - Baseline Mean Breakoff Voltage | 240 | Mean last shock administered in valid baseline sessions (vol | 65.8 |
| ViSTR-Bench | 52.4 | Accuracy (%; averaged over all 1,340 two-option questions in | 55.3 |
Interactive version: theaggregate.ai/model?slug=mimo-v2-5-non-reasoning · How It Works · Data refreshed daily, snapshot 2026-11-05.