Xiaomi · model
MiMo-V2-Flash benchmark scores
As of 2026-10-07, MiMo-V2-Flash (Xiaomi) has measured scores on 15 benchmarks and estimated scores on 72 more.
| Benchmark | Score | Source |
|---|---|---|
| AA-GPQA Diamond | 65.6% | measured |
| AA-HLE | 8.6% | measured |
| AA-SciCode | 44.8% | estimated ± 3.2 pp, high confidence |
| AIME26 | 93.6% | estimated ± 4.6 pp, medium confidence |
| Artificial Analysis Intelligence Index | 16.1% | measured |
| CritPt | 0.0% | measured |
| GDPval-AA | 6.6% | measured |
| GPQA Diamond | 83.3% | estimated ± 0.1 pp, high confidence |
| HLE | 24.4% | estimated ± 8.2 pp, medium confidence |
| IFBench | 55.9% | estimated ± 6.9 pp, low confidence |
| AA-LCR | 36.3% | measured |
| LiveCodeBench v6 | 85.4% | estimated ± 3.1 pp, high confidence |
| AA-Omniscience Accuracy | 15.6% | measured |
| SciCode | 45.4% | estimated ± 3.8 pp, high confidence |
| BrowseComp | 32.9% | estimated ± 10.1 pp, low confidence |
| HLE w/ tools | 35.0% | estimated ± 6.1 pp, low confidence |
| IFEval | 88.5% | estimated ± 1.3 pp, high confidence |
| LongBench v2 | 0.0% | estimated ± 5.9 pp, low confidence |
| τ²-bench results | 83.9% | measured |
| AA Coding Index | 49.8% | measured |
| APEX-Agents | 6.6% | estimated ± 10.0 pp, low confidence |
| APEX-Agents-AA | 9.0% | estimated ± 9.1 pp, medium confidence |
| BioMysteryBench (human-difficult) | 22.7% | estimated ± 5.9 pp, low confidence |
| DeepSearchQA | 53.0% | estimated ± 8.4 pp, medium confidence |
| SWE-bench Verified | 73.4% | measured |
| AutomationBench | 28.4% | estimated ± 9.7 pp, low confidence |
| BFCL v4 | 65.1% | estimated ± 2.6 pp, high confidence |
| CyberGym | 17.1% | estimated ± 6.2 pp, low confidence |
| JobBench | 0.0% | estimated ± 8.6 pp, low confidence |
| SWE-bench Pro | 48.7% | estimated ± 3.2 pp, high confidence |
| WideResearch | 51.9% | estimated ± 9.1 pp, low confidence |
| AA Agentic Index | 4.4% | estimated ± 2.8 pp, high confidence |
| MMLU-Pro | 84.9% | measured |
| GPQA | 83.7% | measured |
| AA-IFBench | 39.9% | measured |
| Gert Labs | 36.9% | estimated ± 8.9 pp, medium confidence |
| AA-AnalystAgent | 4.2% | estimated ± 12.0 pp, low confidence |
| AA EnterpriseOps-Gym | 28.1% | estimated ± 3.6 pp, medium confidence |
| ARC-AGI-1 | 23.9% | estimated ± 5.4 pp, low confidence |
| ARC-AGI-2 | 12.3% | estimated ± 11.6 pp, low confidence |
| CursorBench 3.1 | 47.9% | estimated ± 4.8 pp, low confidence |
| CursorBench 3.2 | 34.7% | estimated ± 4.4 pp, medium confidence |
| FrontierCode 1.1 Main | 7.7% | estimated ± 2.7 pp, low confidence |
| FrontierSWE v2 | 3.7% | estimated ± 14.8 pp, low confidence |
| OSWorld-Verified | 59.0% | estimated ± 8.7 pp, low confidence |
| Vals GPQA Diamond | 59.3% | measured |
| Vals LiveCodeBench | 73.6% | estimated ± 9.9 pp, low confidence |
| Vals MMLU-Pro | 79.5% | estimated ± 1.2 pp, high confidence |
| Vals SWE-bench | 70.2% | estimated ± 5.4 pp, medium confidence |
| AA AutomationBench | 3.7% | estimated ± 7.0 pp, low confidence |
| GDP.pdf | 9.4% | estimated ± 5.4 pp, low confidence |
| AA ITBench | 54.2% | estimated ± 3.6 pp, low confidence |
| AA Tau3 Banking | 22.0% | estimated ± 5.8 pp, low confidence |
| CursorBench 4.0 | 15.9% | estimated ± 1.6 pp, medium confidence |
| CWE-bench v1 | 11.3% | estimated ± 6.9 pp, low confidence |
| DeepSWE | 13.9% | estimated ± 6.4 pp, low confidence |
| HLE w/o tools | 17.9% | estimated ± 4.9 pp, high confidence |
| PostTrainBench v1.1 | 10.8% | estimated ± 6.1 pp, low confidence |
| Toolathlon-Verified | 18.5% | estimated ± 2.4 pp, medium confidence |
| Vibe Code Bench | 1.4% | estimated ± 8.1 pp, medium confidence |
| C-Eval | 89.5% | estimated ± 0.9 pp, low confidence |
| MCP Atlas | 54.2% | estimated ± 11.1 pp, low confidence |
| MCP-Tasks | 73.4% | estimated ± 8.2 pp, low confidence |
| MMLU-Redux | 93.0% | estimated ± 0.7 pp, high confidence |
| NL2Repo | 36.0% | estimated ± 5.1 pp, medium confidence |
| QwenClawBench | 55.9% | estimated ± 6.0 pp, low confidence |
| SuperGPQA | 57.4% | estimated ± 7.5 pp, medium confidence |
| Toolathlon | 26.9% | estimated ± 6.3 pp, low confidence |
| HealthBench Hard | 0.0% | estimated ± 5.1 pp, low confidence |
| MMLU-Pro (Arcee) | 77.1% | estimated ± 4.0 pp, low confidence |
| React Native Evals | 76.6% | estimated ± 3.5 pp, high confidence |
| ResearchClawBench | 8.5% | estimated ± 2.2 pp, low confidence |
| SWE-Rebench | 57.0% | estimated ± 3.6 pp, medium confidence |
| ARC-AGI-3 | 0.0% | estimated ± 2.1 pp, medium confidence |
| BioMysteryBench (human-solvable) | 81.7% | estimated ± 0.6 pp, low confidence |
| DRACO | 64.3% | estimated ± 3.0 pp, low confidence |
| FrontierCode 1.1 Extended | 43.9% | estimated ± 3.4 pp, low confidence |
| HealthBench (raw) | 13.1% | estimated ± 6.4 pp, low confidence |
| HealthBench (length-adjusted) | 53.9% | estimated ± 3.7 pp, low confidence |
| HealthBench Professional | 42.0% | estimated ± 1.7 pp, low confidence |
| HealthBench Professional (raw) | 26.6% | estimated ± 5.0 pp, low confidence |
| HLE-Verified | 15.3% | estimated ± 2.0 pp, low confidence |
| LABBench2 | 46.0% | estimated ± 1.8 pp, low confidence |
| AIME 2025 | 94.1% | measured |
| LiveCodeBench | 98.4% | estimated ± 8.9 pp, low confidence |
| OpenHarmony Bench | 47.4% | estimated ± 2.8 pp, medium confidence |
| MMLU | 82.6% | estimated ± 1.1 pp, medium confidence |