Gemini 2.5 Flash (Sep '25): benchmark results
Provider: Google. Released 2025-09-25. Access: API.
Aggregate Rating 1703 ± 24, rank #334 of 10916 rated models, from 43 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA MMLU-Pro | 84.24 | Accuracy (%) | 88.5 |
| AA Omniscience - Software Engineering (SWE) - Dart | 32 | Accuracy (%) | 87.4 |
| AA Global-MMLU-Lite - Burmese | 86.75 | Accuracy (%) | 87.3 |
| AA Global-MMLU-Lite - Yoruba | 78.33 | Accuracy (%) | 86.4 |
| AA Omniscience - Software Engineering (SWE) - Java | 26 | Accuracy (%) | 86 |
| AA Global-MMLU-Lite - Arabic | 89.33 | Accuracy (%) | 84.5 |
| AA Global-MMLU-Lite - Indonesian | 90.75 | Accuracy (%) | 84.1 |
| AA Global-MMLU-Lite - Swahili | 86.58 | Accuracy (%) | 83.6 |
| AA Omniscience - Software Engineering (SWE) - R | 24 | Accuracy (%) | 83.4 |
| AA Global-MMLU-Lite - Spanish | 91.92 | Accuracy (%) | 83.2 |
| AA Omniscience - Software Engineering (SWE) - HTML | 46 | Accuracy (%) | 83.1 |
| AA Omniscience - Software Engineering (SWE) - Swift | 48 | Accuracy (%) | 82.4 |
Interactive version: theaggregate.ai/model?slug=gemini-2-5-flash-sep-25 · How It Works · Data refreshed daily, snapshot 2026-11-08.