Xiaomi · model
MiMo-V2.5 benchmark scores
As of 2026-11-22, MiMo-V2.5 (Xiaomi) has measured scores on 13 benchmarks and estimated scores on 105 more.
| Benchmark | Score | Source |
|---|---|---|
| AA-GPQA Diamond | 82.5% | estimated ± 3.6 pp, high confidence |
| AA-HLE | 20.2% | estimated ± 6.2 pp, medium confidence |
| AA-SciCode | 48.3% | estimated ± 3.6 pp, high confidence |
| Artificial Analysis Intelligence Index | 19.6% | estimated ± 5.8 pp, medium confidence |
| CritPt | 2.0% | estimated ± 5.3 pp, medium confidence |
| GDPval-AA | 32.8% | estimated ± 8.0 pp, low confidence |
| GPQA Diamond | 86.0% | estimated ± 1.8 pp, medium confidence |
| HLE | 32.8% | estimated ± 9.2 pp, low confidence |
| LiveCodeBench v6 | 87.1% | estimated ± 2.2 pp, medium confidence |
| AA-Omniscience Accuracy | 28.9% | estimated ± 6.5 pp, medium confidence |
| SciCode | 39.7% | estimated ± 2.9 pp, low confidence |
| Terminal-Bench 2.1 | 78.5% | estimated ± 6.7 pp, medium confidence |
| BrowseComp | 81.3% | estimated ± 5.6 pp, medium confidence |
| HLE w/ tools | 48.5% | estimated ± 5.7 pp, medium confidence |
| VITA-Bench | 39.6% | estimated ± 10.7 pp, low confidence |
| AA Coding Index | 53.9% | estimated ± 5.6 pp, medium confidence |
| AA-MMMU-Pro | 77.0% | estimated ± 1.5 pp, high confidence |
| APEX-Agents-AA | 14.7% | estimated ± 9.3 pp, medium confidence |
| DeepSearchQA | 80.5% | estimated ± 10.2 pp, low confidence |
| SWE-bench Verified | 77.6% | estimated ± 3.6 pp, high confidence |
| CyberGym | 54.2% | estimated ± 11.0 pp, low confidence |
| JobBench | 32.5% | estimated ± 9.3 pp, low confidence |
| SWE-bench Pro | 56.1% | measured |
| WideResearch | 70.9% | estimated ± 3.3 pp, high confidence |
| AA Agentic Index | 23.6% | estimated ± 6.4 pp, medium confidence |
| MMLU-Pro | 83.7% | estimated ± 4.8 pp, high confidence |
| GPQA | 83.7% | estimated ± 4.5 pp, high confidence |
| MMLU | 91.2% | estimated ± 1.2 pp, high confidence |
| Vals GPQA Diamond | 81.6% | measured |
| Vals LiveCodeBench | 81.5% | measured |
| Vals MMLU-Pro | 82.9% | measured |
| MMMU | 84.9% | estimated ± 1.7 pp, high confidence |
| AA MMLU-Pro | 85.6% | estimated ± 1.1 pp, high confidence |
| Gert Labs | 29.9% | measured |
| AA Terminal-Bench 2.1 | 80.8% | estimated ± 2.5 pp, high confidence |
| ApprenticeBench | 3.3% | estimated ± 10.4 pp, low confidence |
| ARC-AGI-1 | 71.1% | estimated ± 6.5 pp, medium confidence |
| ARC-AGI-2 | 22.4% | estimated ± 13.8 pp, low confidence |
| CursorBench 3.1 | 52.0% | estimated ± 5.6 pp, low confidence |
| CursorBench 3.2 | 52.3% | estimated ± 3.6 pp, high confidence |
| Terminal-Bench 3.0 | 6.1% | estimated ± 6.9 pp, low confidence |
| FrontierCode 1.1 Main | 37.6% | estimated ± 5.0 pp, high confidence |
| FrontierSWE v2 | 8.8% | estimated ± 8.3 pp, medium confidence |
| MazeBench | 0.0% | estimated ± 2.7 pp, high confidence |
| OSWorld-Verified | 70.4% | estimated ± 3.8 pp, medium confidence |
| ParseBench-1 | 61.2% | estimated ± 8.4 pp, medium confidence |
| Terminal-Bench 4.1 | 12.2% | estimated ± 14.9 pp, low confidence |
| Terminal-Bench Hard | 48.1% | estimated ± 3.9 pp, high confidence |
| Vals SWE-bench | 71.0% | measured |
| Terminal-Bench 2.1 (Vals AI) | 60.7% | measured |
| VulcanBench v3 | 81.6% | estimated ± 5.6 pp, low confidence |
| CWE-Bench-AA | 40.0% | estimated ± 8.1 pp, medium confidence |
| AA Cyber Index | 20.3% | estimated ± 12.4 pp, low confidence |
| DeepsecBench-AA | 9.2% | estimated ± 8.2 pp, low confidence |
| AA ITBench | 36.8% | estimated ± 4.6 pp, medium confidence |
| Bug Hunt Bench | 29.5% | estimated ± 12.1 pp, low confidence |
| CursorBench 4.0 | 32.6% | estimated ± 3.3 pp, medium confidence |
| CWE-bench v1 | 39.3% | estimated ± 5.7 pp, medium confidence |
| DeepSWE | 56.3% | estimated ± 7.2 pp, low confidence |
| FrontierCode 2.0 | 20.1% | estimated ± 6.7 pp, low confidence |
| HLE w/o tools | 26.5% | estimated ± 4.9 pp, high confidence |
| OSWorld 2.0 | 4.6% | estimated ± 6.9 pp, low confidence |
| OSWorld 2.1 | 11.5% | estimated ± 3.6 pp, medium confidence |
| OSWorld 2.1 Offline | 39.9% | estimated ± 10.1 pp, low confidence |
| PostTrainBench v1.1 | 16.2% | estimated ± 3.7 pp, medium confidence |
| Toolathlon-Verified | 63.3% | estimated ± 3.0 pp, high confidence |
| VulcanBench Frontier v4 | 77.4% | estimated ± 8.3 pp, low confidence |
| AA AutomationBench | 43.2% | estimated ± 9.9 pp, low confidence |
| GDP.pdf | 13.4% | estimated ± 4.2 pp, medium confidence |
| AA Terminal-Bench 4.0 | 10.2% | estimated ± 13.7 pp, low confidence |
| ARC-AGI-3 | 0.0% | estimated ± 8.6 pp, low confidence |
| CWE-bench | 33.3% | estimated ± 6.2 pp, low confidence |
| MKA | 32.9% | estimated ± 4.6 pp, medium confidence |
| Hermes Index | 20.4% | estimated ± 5.7 pp, low confidence |
| ParseBench-2 | 69.0% | estimated ± 5.0 pp, medium confidence |
| CharXiv | 81.0% | measured |
| CharXiv w/o tools | 82.0% | estimated ± 1.2 pp, high confidence |
| ExploitGym | 0.0% | estimated ± 8.3 pp, low confidence |
| AI2D_TEST | 93.4% | estimated ± 0.9 pp, medium confidence |
| Claw-Eval | 62.3% | measured |
| CountBench | 96.7% | estimated ± 2.0 pp, medium confidence |
| DynaMath | 86.0% | estimated ± 0.8 pp, medium confidence |
| ERQA | 66.7% | estimated ± 2.9 pp, high confidence |
| MathVision | 87.2% | estimated ± 3.0 pp, high confidence |
| MCP Atlas | 64.6% | estimated ± 12.9 pp, low confidence |
| MCP-Tasks | 77.5% | estimated ± 6.4 pp, low confidence |
| MLVU (M-Avg) | 86.0% | estimated ± 1.5 pp, high confidence |
| MMMU-Pro | 77.9% | measured |
| NL2Repo | 40.5% | estimated ± 4.7 pp, high confidence |
| PostTrainBench v1 | 13.2% | estimated ± 3.6 pp, medium confidence |
| RealWorldQA | 83.8% | estimated ± 1.5 pp, high confidence |
| ScreenSpot Pro | 73.2% | estimated ± 6.8 pp, medium confidence |
| SimpleVQA | 66.2% | estimated ± 7.5 pp, medium confidence |
| Terminal-Bench 2.0 | 65.8% | measured |
| Toolathlon | 52.3% | estimated ± 6.9 pp, medium confidence |
| Video-MME (w/o subtitle) | 84.4% | estimated ± 1.3 pp, high confidence |
| Video-MME (with subtitle) | 87.7% | measured |
| VideoMMMU | 85.4% | estimated ± 1.5 pp, medium confidence |
| We-Math | 86.6% | estimated ± 3.5 pp, medium confidence |
| AA LiveCodeBench | 87.1% | estimated ± 2.6 pp, high confidence |
| LiveCodeBench Pro | 80.6% | estimated ± 6.0 pp, low confidence |
| MedXpertQA (MM) | 70.0% | estimated ± 2.4 pp, high confidence |
| React Native Evals | 77.3% | estimated ± 4.7 pp, medium confidence |
| ResearchClawBench | 16.9% | measured |
| PostTrainBench | 13.7% | estimated ± 3.7 pp, medium confidence |
| Finance Agent v2 | 55.1% | estimated ± 2.4 pp, low confidence |
| FrontierCode 1.1 Extended | 54.7% | estimated ± 3.0 pp, low confidence |
| LABBench2 | 79.1% | estimated ± 1.8 pp, low confidence |
| ScarfBench | 42.1% | estimated ± 12.3 pp, low confidence |
| Agents' Last Exam | 29.7% | estimated ± 9.7 pp, low confidence |
| BabyVision w/ Python | 88.8% | estimated ± 3.8 pp, low confidence |
| OpenHarmony Bench v1.0 | 51.5% | estimated ± 3.2 pp, high confidence |
| BabyVision | 60.5% | estimated ± 9.1 pp, low confidence |
| MMVU | 77.1% | estimated ± 2.7 pp, medium confidence |
| PerceptionBench | 59.6% | estimated ± 4.9 pp, low confidence |
| OCRBench V2 | 71.1% | estimated ± 3.3 pp, high confidence |
| LifeVal | 6.9% | estimated ± 5.0 pp, low confidence |
| MMMU-Pro w/ Python | 80.5% | estimated ± 2.6 pp, high confidence |