Google · model
Gemini 3.7 Flash benchmark scores
As of 2026-10-07, Gemini 3.7 Flash (Google) has measured scores on 35 benchmarks and estimated scores on 76 more.
| Benchmark | Score | Source |
|---|---|---|
| AA-GPQA Diamond | 94.5% | measured |
| AA-HLE | 47.9% | measured |
| AA-SciCode | 57.2% | measured |
| Artificial Analysis Intelligence Index | 39.1% | measured |
| CritPt | 14.3% | measured |
| GDPval-AA | 44.6% | measured |
| GPQA Diamond | 93.8% | estimated ± 1.4 pp, high confidence |
| HLE | 57.8% | estimated ± 6.3 pp, medium confidence |
| AA-LCR | 81.7% | measured |
| LiveCodeBench v6 | 93.4% | estimated ± 3.1 pp, medium confidence |
| AA-Omniscience Accuracy | 55.3% | measured |
| SciCode | 54.2% | estimated ± 2.9 pp, low confidence |
| Terminal-Bench 2.1 | 85.8% | measured |
| BrowseComp | 86.1% | estimated ± 2.2 pp, high confidence |
| HLE w/ tools | 60.2% | estimated ± 5.7 pp, low confidence |
| LongBench v2 | 61.5% | estimated ± 5.9 pp, low confidence |
| τ²-bench results | 94.3% | estimated ± 7.1 pp, low confidence |
| AA Coding Index | 76.1% | measured |
| AA-MMMU-Pro | 85.5% | measured |
| APEX-Agents | 37.5% | estimated ± 10.0 pp, low confidence |
| APEX-Agents-AA | 38.4% | estimated ± 2.9 pp, medium confidence |
| BioMysteryBench (human-difficult) | 43.5% | measured |
| DeepSearchQA | 90.9% | estimated ± 4.5 pp, high confidence |
| SWE-bench Verified | 86.2% | estimated ± 0.8 pp, medium confidence |
| AutomationBench | 30.4% | measured |
| BFCL v4 | 77.0% | estimated ± 2.6 pp, medium confidence |
| CyberGym | 79.6% | estimated ± 3.7 pp, high confidence |
| JobBench | 62.9% | estimated ± 4.9 pp, medium confidence |
| SWE-bench Pro | 74.1% | estimated ± 4.8 pp, high confidence |
| WideResearch | 95.1% | estimated ± 9.1 pp, low confidence |
| AA Agentic Index | 36.4% | measured |
| MMLU-Pro | 85.7% | estimated ± 3.5 pp, medium confidence |
| GPQA | 93.7% | estimated ± 1.6 pp, medium confidence |
| Gert Labs | 69.7% | estimated ± 3.0 pp, medium confidence |
| AA-AnalystAgent | 60.0% | measured |
| AA EnterpriseOps-Gym | 57.6% | estimated ± 1.8 pp, low confidence |
| AA Harvey LAB | 90.7% | measured |
| ApprenticeBench | 16.0% | measured |
| ARC-AGI-1 | 95.5% | measured |
| ARC-AGI-2 | 84.6% | measured |
| CursorBench 3.1 | 61.8% | estimated ± 3.4 pp, medium confidence |
| CursorBench 3.2 | 60.7% | estimated ± 1.6 pp, medium confidence |
| Terminal-Bench 3.0 | 14.9% | measured |
| FrontierCode 1.1 Main | 43.6% | measured |
| FrontierSWE v2 | 20.3% | measured |
| OfficeQA Pro | 67.7% | estimated ± 7.2 pp, medium confidence |
| OSWorld-Verified | 78.0% | estimated ± 1.9 pp, medium confidence |
| Vals GPQA Diamond | 93.9% | measured |
| Vals LiveCodeBench | 88.7% | measured |
| Vals MMLU-Pro | 90.1% | measured |
| Vals SWE-bench | 80.8% | measured |
| Terminal-Bench 2.1 (Vals AI) | 77.5% | measured |
| VulcanBench v3 | 86.3% | estimated ± 5.6 pp, low confidence |
| AA AutomationBench | 49.7% | estimated ± 7.0 pp, medium confidence |
| GDP.pdf | 13.3% | estimated ± 3.4 pp, high confidence |
| AA ITBench | 44.0% | estimated ± 6.6 pp, low confidence |
| AA Tau3 Banking | 47.7% | estimated ± 4.4 pp, low confidence |
| AA Terminal-Bench 2.1 | 84.9% | estimated ± 3.5 pp, high confidence |
| AA Terminal-Bench 4.0 | 25.7% | estimated ± 8.9 pp, medium confidence |
| CursorBench 4.0 | 41.1% | estimated ± 1.6 pp, high confidence |
| CWE-bench v1 | 52.1% | estimated ± 6.9 pp, medium confidence |
| DeepSWE | 65.3% | measured |
| HLE w/o tools | 51.9% | estimated ± 5.1 pp, medium confidence |
| OSWorld 2.0 | 47.9% | measured |
| PostTrainBench v1.1 | 24.3% | estimated ± 2.7 pp, high confidence |
| Toolathlon-Verified | 78.5% | estimated ± 1.0 pp, medium confidence |
| Vibe Code Bench | 72.6% | estimated ± 5.6 pp, low confidence |
| CharXiv | 88.7% | measured |
| CharXiv w/o tools | 84.5% | measured |
| ExploitGym | 18.0% | estimated ± 6.9 pp, low confidence |
| C-Eval | 95.5% | estimated ± 1.0 pp, low confidence |
| Claw-Eval | 77.6% | estimated ± 13.6 pp, low confidence |
| MathVision | 91.5% | estimated ± 2.1 pp, medium confidence |
| MCP Atlas | 84.8% | estimated ± 6.2 pp, low confidence |
| MMLU-Redux | 97.6% | estimated ± 1.1 pp, low confidence |
| MMMU-Pro | 81.4% | estimated ± 1.4 pp, medium confidence |
| NL2Repo | 46.4% | estimated ± 4.7 pp, high confidence |
| QwenClawBench | 100.0% | estimated ± 6.0 pp, low confidence |
| ScreenSpot Pro | 84.7% | estimated ± 7.1 pp, medium confidence |
| SuperGPQA | 87.6% | estimated ± 7.5 pp, low confidence |
| Terminal-Bench 2.0 | 78.3% | estimated ± 5.9 pp, low confidence |
| Toolathlon | 54.9% | estimated ± 1.7 pp, medium confidence |
| VideoMMMU | 86.8% | estimated ± 1.5 pp, high confidence |
| ERQA | 70.6% | estimated ± 3.7 pp, high confidence |
| HealthBench Hard | 33.2% | estimated ± 7.7 pp, low confidence |
| MedXpertQA (MM) | 78.5% | estimated ± 6.4 pp, low confidence |
| MMLU-Pro (Arcee) | 100.0% | estimated ± 4.0 pp, low confidence |
| React Native Evals | 82.6% | estimated ± 2.4 pp, medium confidence |
| ResearchClawBench | 19.6% | estimated ± 2.0 pp, high confidence |
| SWE-Rebench | 58.1% | estimated ± 8.9 pp, low confidence |
| ARC-AGI-3 | 1.7% | estimated ± 3.4 pp, high confidence |
| BioMysteryBench (human-solvable) | 87.1% | measured |
| DRACO | 79.8% | estimated ± 3.0 pp, medium confidence |
| FrontierCode 1.1 Extended | 31.7% | estimated ± 2.4 pp, low confidence |
| HealthBench (raw) | 56.1% | estimated ± 6.4 pp, low confidence |
| HealthBench (length-adjusted) | 57.2% | estimated ± 2.5 pp, low confidence |
| HealthBench Professional | 60.0% | estimated ± 1.3 pp, medium confidence |
| HealthBench Professional (raw) | 66.0% | estimated ± 5.0 pp, medium confidence |
| HLE-Verified | 53.6% | measured |
| LABBench2 | 82.1% | measured |
| MMMU | 92.1% | estimated ± 0.9 pp, medium confidence |
| LiveCodeBench | 100.0% | estimated ± 8.9 pp, low confidence |
| Agents' Last Exam | 26.3% | measured |
| OpenHarmony Bench | 58.8% | estimated ± 2.4 pp, high confidence |
| SimpleVQA | 71.2% | estimated ± 7.8 pp, low confidence |
| MMLU | 95.6% | estimated ± 1.1 pp, low confidence |
| MMMU-Pro w/ Python | 85.0% | estimated ± 1.6 pp, medium confidence |
| MathVision w/ Python | 85.5% | estimated ± 0.9 pp, low confidence |
| RealWorldQA | 87.0% | estimated ± 1.1 pp, medium confidence |
| Video-MME (with subtitle) | 89.2% | estimated ± 0.9 pp, medium confidence |
| OmniDocBench 1.5 | 92.0% | estimated ± 6.4 pp, low confidence |