Google
availableShows if the model has enough results for an index.Gemini 3.6 Flash
Gemini 3.6 Flash is a reasoning model from Google. 41 benchmarks count toward its score, in 8 categories.
IndexOverall score. 50 is the middle.66.8 ±2.5
CoverageShare of the index weight with results.100%
SpeedOutput tokens per second.133/s
Input / 1MUS dollars per 1M input tokens.$0.75 batch $0.375
Output / 1MUS dollars per 1M output tokens.$3.75 batch $1.88 US dollars per 1M output tokens in a batch.
ContextMaximum tokens in one request.1.05M
EloLMArena rating and rank.1476 (#18)
50 is the middle of the board. The range shows the doubt in the index. Batch work costs less.
26,445 votes. Elo shows what people prefer. It does not change the score.
CapabilitiesScore per category. 50 is the middle.
50 is the middleResults
41 counted| BenchmarkThe test name. | CategoryThe capability that the test measures. | ResultThe score from the publisher. | IndexThis result on the index scale. | RunThe settings of the run. | DateDate of the result. | Published byThe source of the result. |
|---|---|---|---|---|---|---|
| OTIS Mock AIME 2024-2025 | Math | 94.2% | 64.2 | high effort | — | Epoch AI |
| GPQA diamond | Knowledge | 94.1% | 65.1 | high effort | — | Epoch AI |
| GPQA Diamond | Knowledge | 93.4% | 64.5 | — | 1 Sept 2026 | Vals AI |
| Artificial Analysis GPQA Diamond | Knowledge | 92.8% | 63.7 | — | — | Artificial Analysis |
| ARC-AGI-1 (semi-private) | Reasoning | 91.2% | 70.3 | high effort | — | ARC Prize Foundation |
| MMLU Pro | Knowledge | 89.3% | 61.2 | — | 1 Sept 2026 | Vals AI |
| MMMU Pro | Multimodal | 88.4% | 71.2 | — | 1 Sept 2026 | Vals AI |
| LiveCodeBench | Coding | 88.1% | 64.4 | — | 1 Sept 2026 | Vals AI |
| LiveBench Mathematics | Math | 86.4% | 63.3 | high effort | 25 Jun 2026 | LiveBench |
| LiveBench Reasoning | Reasoning | 85.1% | 74.2 | high effort | 25 Jun 2026 | LiveBench |
| LiveBench Language | Knowledge | 83.9% | 71.8 | high effort | 25 Jun 2026 | LiveBench |
| Artificial Analysis MMMU-Pro | Multimodal | 83.2% | 68.8 | — | — | Artificial Analysis |
| OSWorld-Verified | Agentic | 83.0% | 71.0 | — | — | Tianbao Xie et al. |
| Artificial Analysis Long Context Reasoning | Reasoning | 80.0% | 63.6 | — | — | Artificial Analysis |
| SWE-bench | Coding | 79.6% | 61.4 | — | 1 Sept 2026 | Vals AI |
| LiveBench Coding | Coding | 77.9% | 66.9 | high effort | 25 Jun 2026 | LiveBench |
| EuroEval Swedish | Multilingual | 75.9% | 95.0 | — | — | EuroEval |
| EuroEval French | Multilingual | 75.6% | 95.0 | — | — | EuroEval |
| LiveBench Instruction Following | Instruction | 75.4% | 78.9 | high effort | 25 Jun 2026 | LiveBench |
| EuroEval Italian | Multilingual | 75.1% | 95.0 | — | — | EuroEval |
| Terminal-Bench 2.1 | Agentic | 73.8% | 67.6 | — | 21 Sept 2026 | Vals AI |
| EuroEval Portuguese | Multilingual | 72.6% | 92.9 | — | — | EuroEval |
| EuroEval Dutch | Multilingual | 71.8% | 91.9 | — | — | EuroEval |
| Artificial Analysis Coding Index | Coding | 69.2% | 67.7 | — | — | Artificial Analysis |
| EuroEval Spanish | Multilingual | 66.8% | 85.7 | — | — | EuroEval |
| SimpleQA Verified | Knowledge | 66.2% | 82.8 | high effort | — | Epoch AI |
| EuroEval German | Multilingual | 66.0% | 84.6 | — | — | EuroEval |
| EuroEval Polish | Multilingual | 65.9% | 84.5 | — | — | EuroEval |
| Vibe Code Bench v1.1 | Coding | 64.0% | 68.8 | OpenHands | 21 Sept 2026 | Vals AI |
| LiveBench Data Analysis | Reasoning | 63.0% | 43.4 | high effort | 25 Jun 2026 | LiveBench |
| ARC-AGI-2 (semi-private) | Reasoning | 60.4% | 70.2 | high effort | — | ARC Prize Foundation |
| FrontierMath-Tiers-1-3-v2-Private | Math | 58.9% | 64.6 | high effort | — | Epoch AI |
| cursorBench32 | Coding | 53.5% | 62.4 | — | — | Benchmark authors |
| Artificial Analysis SciCode | Coding | 53.4% | 67.0 | — | — | Artificial Analysis |
| Artificial Analysis Omniscience Accuracy | Knowledge | 50.0% | 75.7 | — | — | Artificial Analysis |
| DeepSWE | Agentic | 49.0% | 59.4 | — | — | Datacurve AI |
| LiveBench Agentic Coding | Agentic | 43.4% | 57.6 | high effort | 25 Jun 2026 | LiveBench |
| Artificial Analysis Humanity's Last Exam | Knowledge | 40.8% | 68.8 | — | — | Artificial Analysis |
| Chess Puzzles | Reasoning | 40.0% | 75.2 | high effort | — | Epoch AI |
| GDPval-AA normalized | Agentic | 38.2% | 64.9 | — | — | Artificial Analysis |
| IOI | Coding | 35.1% | 60.9 | — | 21 Sept 2026 | Vals AI |
| Artificial Analysis Intelligence Index | Knowledge | 34.0% | 64.9 | — | — | Artificial Analysis |
| Code Migration | Coding | 30.9% | 64.7 | — | 21 Sept 2026 | Vals AI |
| Artificial Analysis Agentic Index | Agentic | 30.1% | 61.8 | — | — | Artificial Analysis |
| Mystery Game Puzzles | Reasoning | 30.0% | 65.8 | high effort | — | Epoch AI |
| Furniture Assembly | Reasoning | 23.3% | 56.1 | high effort | — | Epoch AI |
| FrontierMath-Tier-4-v2-Private | Math | 22.0% | 58.5 | high effort | — | Epoch AI |
| Critical Physics Tasks | Reasoning | 10.6% | 60.6 | — | — | Artificial Analysis |
| Terminal-Bench 4.0 | Agentic | 4.5% | 62.7 | — | 21 Sept 2026 | Vals AI |
| ProgramBench | Coding | 0.0% | — | — | 21 Sept 2026 | Vals AI |
| Agent Arena steerability | Agentic | -3.4 | 63.6 | high effort | 15 Sept 2026 | LMArena |
| Agent Arena task outcome | Agentic | -6.8 | 59.7 | high effort | 15 Sept 2026 | LMArena |
| Agent Arena command recovery | Agentic | -12.4 | 53.4 | high effort | 15 Sept 2026 | LMArena |
41 benchmarks count, from 52 of 53 results. A grey row does not count. Too few models took that benchmark.