Google · model
Gemini 3.6 Flash benchmark scores
As of 2026-10-07, Gemini 3.6 Flash (Google) has measured scores on 21 benchmarks and estimated scores on 85 more.
| Benchmark | Score | Source |
|---|---|---|
| AA-GPQA Diamond | 92.8% | measured |
| AA-HLE | 40.8% | measured |
| AA-SciCode | 53.4% | measured |
| Artificial Analysis Intelligence Index | 34.0% | measured |
| CritPt | 10.6% | measured |
| GDPval-AA | 39.3% | measured |
| GPQA Diamond | 91.8% | estimated ± 1.4 pp, high confidence |
| HLE | 46.0% | estimated ± 6.3 pp, medium confidence |
| AA-LCR | 80.0% | measured |
| LiveCodeBench v6 | 91.7% | estimated ± 3.1 pp, high confidence |
| AA-Omniscience Accuracy | 50.0% | measured |
| SciCode | 52.9% | estimated ± 2.9 pp, low confidence |
| Terminal-Bench 2.1 | 85.2% | estimated ± 6.9 pp, medium confidence |
| BrowseComp | 83.8% | estimated ± 3.1 pp, high confidence |
| HLE w/ tools | 51.8% | estimated ± 6.1 pp, medium confidence |
| LongBench v2 | 61.5% | estimated ± 5.9 pp, low confidence |
| τ²-bench results | 93.0% | estimated ± 13.3 pp, low confidence |
| AA Coding Index | 69.2% | measured |
| AA-MMMU-Pro | 83.2% | measured |
| APEX-Agents | 33.8% | estimated ± 10.0 pp, low confidence |
| APEX-Agents-AA | 30.2% | estimated ± 9.1 pp, medium confidence |
| BioMysteryBench (human-difficult) | 45.6% | estimated ± 5.9 pp, low confidence |
| DeepSearchQA | 89.2% | estimated ± 4.5 pp, high confidence |
| SWE-bench Verified | 89.6% | estimated ± 4.4 pp, high confidence |
| AutomationBench | 28.4% | estimated ± 9.7 pp, low confidence |
| BFCL v4 | 76.4% | estimated ± 2.6 pp, medium confidence |
| CyberGym | 68.7% | estimated ± 5.3 pp, medium confidence |
| JobBench | 38.7% | estimated ± 8.6 pp, medium confidence |
| SWE-bench Pro | 57.0% | estimated ± 4.5 pp, high confidence |
| WideResearch | 90.3% | estimated ± 9.1 pp, low confidence |
| AA Agentic Index | 30.2% | measured |
| MMLU-Pro | 85.7% | estimated ± 3.5 pp, medium confidence |
| GPQA | 93.0% | estimated ± 1.6 pp, high confidence |
| Gert Labs | 76.4% | estimated ± 7.2 pp, medium confidence |
| AA-AnalystAgent | 29.9% | estimated ± 12.0 pp, low confidence |
| AA EnterpriseOps-Gym | 47.9% | estimated ± 5.9 pp, low confidence |
| AA Harvey LAB | 87.7% | estimated ± 1.2 pp, medium confidence |
| ApprenticeBench | 26.7% | estimated ± 5.2 pp, low confidence |
| ARC-AGI-1 | 91.2% | measured |
| ARC-AGI-2 | 60.4% | measured |
| CursorBench 3.1 | 52.4% | estimated ± 3.4 pp, medium confidence |
| CursorBench 3.2 | 53.5% | measured |
| Terminal-Bench 3.0 | 17.7% | estimated ± 6.9 pp, medium confidence |
| FrontierCode 1.1 Main | 36.5% | estimated ± 1.6 pp, low confidence |
| FrontierSWE v2 | 10.1% | estimated ± 9.5 pp, medium confidence |
| OfficeQA Pro | 67.7% | estimated ± 7.2 pp, medium confidence |
| OSWorld-Verified | 83.0% | measured |
| Vals GPQA Diamond | 93.4% | measured |
| Vals LiveCodeBench | 88.1% | measured |
| Vals MMLU-Pro | 89.3% | measured |
| Vals SWE-bench | 79.6% | measured |
| Terminal-Bench 2.1 (Vals AI) | 73.8% | measured |
| VulcanBench v3 | 75.4% | estimated ± 5.1 pp, low confidence |
| AA AutomationBench | 26.2% | estimated ± 7.0 pp, medium confidence |
| GDP.pdf | 9.4% | estimated ± 5.4 pp, medium confidence |
| AA Tau3 Banking | 22.1% | estimated ± 5.8 pp, medium confidence |
| AA Terminal-Bench 2.1 | 86.2% | estimated ± 4.4 pp, high confidence |
| AA Terminal-Bench 4.0 | 29.3% | estimated ± 14.0 pp, low confidence |
| CursorBench 4.0 | 31.7% | estimated ± 1.6 pp, medium confidence |
| CWE-bench v1 | 48.0% | estimated ± 6.9 pp, medium confidence |
| DeepSWE | 49.0% | measured |
| HLE w/o tools | 49.1% | estimated ± 5.1 pp, medium confidence |
| OSWorld 2.0 | 18.1% | estimated ± 3.6 pp, high confidence |
| PostTrainBench v1.1 | 23.8% | estimated ± 2.7 pp, high confidence |
| ProgramBench | 62.2% | estimated ± 6.0 pp, low confidence |
| Toolathlon-Verified | 60.0% | estimated ± 2.3 pp, medium confidence |
| Vibe Code Bench | 50.7% | estimated ± 10.3 pp, low confidence |
| CharXiv | 86.0% | estimated ± 7.5 pp, low confidence |
| CharXiv w/o tools | 84.6% | estimated ± 2.4 pp, high confidence |
| ExploitGym | 4.4% | estimated ± 12.2 pp, low confidence |
| C-Eval | 94.8% | estimated ± 1.0 pp, low confidence |
| Claw-Eval | 73.3% | estimated ± 13.6 pp, low confidence |
| MathVision | 89.7% | estimated ± 4.9 pp, medium confidence |
| MCP Atlas | 78.7% | estimated ± 7.8 pp, low confidence |
| MMLU-Redux | 97.4% | estimated ± 1.1 pp, low confidence |
| MMMU-Pro | 80.6% | estimated ± 4.4 pp, medium confidence |
| NL2Repo | 45.6% | estimated ± 4.7 pp, high confidence |
| QwenClawBench | 100.0% | estimated ± 6.0 pp, low confidence |
| ScreenSpot Pro | 82.7% | estimated ± 10.7 pp, low confidence |
| SuperGPQA | 85.7% | estimated ± 7.5 pp, low confidence |
| Terminal-Bench 2.0 | 76.0% | estimated ± 5.9 pp, medium confidence |
| Toolathlon | 55.8% | estimated ± 6.3 pp, medium confidence |
| ERQA | 68.8% | estimated ± 3.9 pp, medium confidence |
| HealthBench Hard | 30.7% | estimated ± 7.7 pp, low confidence |
| MedXpertQA (MM) | 77.6% | estimated ± 6.5 pp, low confidence |
| MMLU-Pro (Arcee) | 100.0% | estimated ± 4.0 pp, low confidence |
| React Native Evals | 80.6% | estimated ± 2.4 pp, medium confidence |
| ResearchClawBench | 18.7% | 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) | 83.8% | estimated ± 0.6 pp, low confidence |
| DRACO | 77.7% | estimated ± 3.0 pp, medium confidence |
| FrontierCode 1.1 Extended | 55.1% | estimated ± 1.6 pp, low confidence |
| HealthBench (raw) | 49.8% | estimated ± 6.4 pp, low confidence |
| HealthBench (length-adjusted) | 55.6% | estimated ± 2.5 pp, low confidence |
| HealthBench Professional | 59.9% | estimated ± 1.3 pp, medium confidence |
| HealthBench Professional (raw) | 60.1% | estimated ± 5.0 pp, medium confidence |
| HLE-Verified | 35.3% | estimated ± 0.7 pp, low confidence |
| LABBench2 | 73.6% | estimated ± 1.3 pp, low confidence |
| MMMU | 90.1% | estimated ± 0.9 pp, medium confidence |
| LiveCodeBench | 100.0% | estimated ± 8.9 pp, low confidence |
| OpenHarmony Bench | 58.7% | estimated ± 2.4 pp, high confidence |
| MMLU | 94.9% | estimated ± 1.1 pp, low confidence |
| MMMU-Pro w/ Python | 83.7% | estimated ± 1.6 pp, high confidence |
| RealWorldQA | 89.2% | estimated ± 1.2 pp, low confidence |
| OmniDocBench 1.5 | 92.0% | estimated ± 6.4 pp, low confidence |