Gemini 2.5 Flash (Preview 2504): benchmark results
Provider: Google. Access: API.
Aggregate Rating 1638 ± 28, rank #702 of 10916 rated models, from 20 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AGI-Eval Community - Learning (Chinese) | 84.11 | Accuracy (%) | 85.5 |
| AGI-Eval Community - Learning | 85.9 | Accuracy (%) | 59.3 |
| AGI-Eval Community - Cognition (Chinese) | 77.27 | Accuracy (%) | 55.8 |
| AGI-Eval Community - Cognition | 77.63 | Accuracy (%) | 50 |
| AGI-Eval Community - Instruction Following | 37.59 | Accuracy (%) | 50 |
| AGI-Eval Community - Mathematical Reasoning | 70.5 | Accuracy (%) | 43.6 |
| AGI-Eval Community - Cognition (English) | 78.14 | Accuracy (%) | 42 |
| AGI-Eval Community - Algorithmic Reasoning | 28.13 | Accuracy (%) | 40.7 |
| AGI-Eval Community - Objective Accuracy (Chinese) | 56.54 | Accuracy (%) | 39.9 |
| AGI-Eval Community - General Reasoning | 95.42 | Accuracy (%) | 39.6 |
| AGI-Eval Community - Subject Knowledge | 81.26 | Accuracy (%) | 38.6 |
| AGI-Eval Community - Interaction (English) | 88.94 | Accuracy (%) | 37 |
Interactive version: theaggregate.ai/model?slug=gemini-2-5-flash-preview-2504 · How It Works · Data refreshed daily, snapshot 2026-11-08.