Google
availableShows if the model has enough results for an index.Gemini 3.5 Flash-Lite
Gemini 3.5 Flash-Lite is a reasoning model from Google. 38 benchmarks count toward its score, in 8 categories.
IndexOverall score. 50 is the middle.54.4 ±2.6
CoverageShare of the index weight with results.100%
SpeedOutput tokens per second.73/s
Input / 1MUS dollars per 1M input tokens.$0.3
Output / 1MUS dollars per 1M output tokens.$2.5
ContextMaximum tokens in one request.1.05M
EloLMArena rating and rank.1436 (#79)
50 is the middle of the board. The range shows the doubt in the index.
26,165 votes. Elo shows what people prefer. It does not change the score.
CapabilitiesScore per category. 50 is the middle.
50 is the middleResults
38 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. |
|---|---|---|---|---|---|---|
| MMLU Pro | Knowledge | 85.8% | 55.8 | — | 1 Sept 2026 | Vals AI |
| GPQA Diamond | Knowledge | 83.8% | 55.6 | — | 1 Sept 2026 | Vals AI |
| Artificial Analysis GPQA Diamond | Knowledge | 83.8% | 54.4 | — | — | Artificial Analysis |
| MMMU Pro | Multimodal | 83.6% | 63.4 | — | 1 Sept 2026 | Vals AI |
| GPQA diamond | Knowledge | 83.3% | 55.2 | high effort | — | Epoch AI |
| LiveCodeBench | Coding | 79.0% | 56.2 | — | 1 Sept 2026 | Vals AI |
| Artificial Analysis MMMU-Pro | Multimodal | 79.0% | 63.6 | — | — | Artificial Analysis |
| LiveBench Coding | Coding | 76.1% | 64.0 | high effort | 25 Jun 2026 | LiveBench |
| Artificial Analysis Long Context Reasoning | Reasoning | 76.0% | 60.8 | — | — | Artificial Analysis |
| SWE-bench | Coding | 75.0% | 57.7 | — | 1 Sept 2026 | Vals AI |
| OSWorld-Verified | Agentic | 74.0% | 62.6 | — | — | Tianbao Xie et al. |
| LiveBench Mathematics | Math | 73.7% | 46.3 | high effort | 25 Jun 2026 | LiveBench |
| MRCRv2 | Reasoning | 72.2% | — | — | — | OpenAI |
| LiveBench Language | Knowledge | 71.8% | 57.3 | high effort | 25 Jun 2026 | LiveBench |
| OTIS Mock AIME 2024-2025 | Math | 71.1% | 51.3 | high effort | — | Epoch AI |
| LiveBench Instruction Following | Instruction | 67.2% | 66.2 | high effort | 25 Jun 2026 | LiveBench |
| EuroEval Swedish | Multilingual | 63.0% | 81.0 | — | — | EuroEval |
| EuroEval French | Multilingual | 61.6% | 79.2 | — | — | EuroEval |
| EuroEval Italian | Multilingual | 61.6% | 79.1 | — | — | EuroEval |
| EuroEval Portuguese | Multilingual | 61.3% | 78.8 | — | — | EuroEval |
| LiveBench Reasoning | Reasoning | 60.2% | 39.5 | high effort | 25 Jun 2026 | LiveBench |
| EuroEval Dutch | Multilingual | 59.6% | 76.7 | — | — | EuroEval |
| EuroEval Spanish | Multilingual | 57.1% | 73.6 | — | — | EuroEval |
| EuroEval Polish | Multilingual | 54.3% | 70.1 | — | — | EuroEval |
| SWE-bench Pro | Coding | 54.2% | 56.4 | — | — | Xiang Deng et al. |
| ARC-AGI-1 (semi-private) | Reasoning | 53.5% | 52.5 | high effort | — | ARC Prize Foundation |
| LiveBench Data Analysis | Reasoning | 53.2% | 29.9 | high effort | 25 Jun 2026 | LiveBench |
| EuroEval German | Multilingual | 52.4% | 67.8 | — | — | EuroEval |
| Terminal-Bench 2.1 | Agentic | 50.2% | 53.7 | — | 21 Sept 2026 | Vals AI |
| Artificial Analysis Coding Index | Coding | 49.3% | 53.7 | — | — | Artificial Analysis |
| LiveBench Agentic Coding | Agentic | 45.3% | 59.3 | high effort | 25 Jun 2026 | LiveBench |
| Artificial Analysis EnterpriseOps-Gym | Agentic | 42.3% | 65.3 | — | — | Artificial Analysis |
| Artificial Analysis SciCode | Coding | 41.3% | 50.2 | — | — | Artificial Analysis |
| Vibe Code Bench v1.1 | Coding | 37.2% | 57.6 | OpenHands | 21 Sept 2026 | Vals AI |
| Artificial Analysis Omniscience Accuracy | Knowledge | 29.5% | 50.4 | — | — | Artificial Analysis |
| IOI v1 | Coding | 26.2% | 54.8 | — | 9 Aug 2026 | Vals AI |
| FrontierMath-Tiers-1-3-v2-Private | Math | 26.0% | 46.1 | high effort | — | Epoch AI |
| GDPval-AA normalized | Agentic | 23.5% | 53.6 | — | — | Artificial Analysis |
| Artificial Analysis Intelligence Index | Knowledge | 22.2% | 50.1 | — | — | Artificial Analysis |
| Chess Puzzles | Reasoning | 22.0% | 52.0 | high effort | — | Epoch AI |
| Mystery Game Puzzles | Reasoning | 19.0% | 54.1 | low effort | — | Epoch AI |
| Artificial Analysis Humanity's Last Exam | Knowledge | 18.8% | 44.9 | — | — | Artificial Analysis |
| Artificial Analysis Agentic Index | Agentic | 15.9% | 50.2 | — | — | Artificial Analysis |
| ARC-AGI-2 (semi-private) | Reasoning | 10.3% | 44.9 | high effort | — | ARC Prize Foundation |
| Code Migration | Coding | 6.1% | 48.8 | — | 21 Sept 2026 | Vals AI |
| Critical Physics Tasks | Reasoning | 0.0% | 38.4 | — | — | Artificial Analysis |
| ProgramBench | Coding | 0.0% | — | — | 21 Sept 2026 | Vals AI |
| FrontierMath-Tier-4-v2-Private | Math | 0.0% | 47.9 | high effort | — | Epoch AI |
| Agent Arena steerability | Agentic | -12.8 | 53.0 | — | 15 Sept 2026 | LMArena |
| Agent Arena task outcome | Agentic | -17.6 | 47.5 | — | 15 Sept 2026 | LMArena |
| Agent Arena command recovery | Agentic | -28.7 | 35.0 | — | 15 Sept 2026 | LMArena |
38 benchmarks count, from 49 of 51 results. A grey row does not count. Too few models took that benchmark.