Gemini 2.0 Flash (Thinking): benchmark results
Gemini 2.0 Flash evaluated with thinking enabled. Provider: Google. Released 2024-12-11. Access: API.
Aggregate Rating 1598 ± 12, rank #722 of 2168 rated models, from 90 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| WildToolBench | 14.45 | Session accuracy (%): share of the 256 dialogues in which al | 100 |
| WildToolBench - Clarification Tasks | 52.34 | Accuracy (%) on tasks where the correct response is to ask t | 100 |
| WildToolBench - Task Accuracy | 61.04 | Task accuracy (%) over all 1,024 tasks, on WildToolBench (25 | 100 |
| MMVU | 69.5 | Hidden-test overall accuracy, chain-of-thought (%) | 97.1 |
| Video-Holmes | 43.1 | Overall (imported) | 93.3 |
| TimeSpot - Continent | 88.66 | Accuracy (%) of the predicted continent on TimeSpot's 1,455 | 88.5 |
| TimeSpot - Country | 76.22 | Accuracy (%) of the predicted country on TimeSpot's 1,455 gr | 88.5 |
| WildToolBench - Multi-Tool Tasks | 40.23 | Accuracy (%) on compositional tasks needing multi-step seque | 86.6 |
| SEAL - Arabic | 1120 | Score | 86.1 |
| SpatialThinker Release - 3DSRLocation | 67.1 | Accuracy (%) | 85.7 |
| SEAL - Spanish | 1108 | Score | 81 |
| AGL1K | 2494.46 | Geo-score (0-5000) | 80 |
Interactive version: theaggregate.ai/model?slug=gemini-2-0-flash-thinking · How It Works · Data refreshed daily, snapshot 2026-10-20.