Gemini 2.5 Flash (Preview 09-2025) — benchmark results
September 2025 Gemini 2.5 Flash preview snapshot. Provider: Google. Released 2025-09-25. Access: API.
Unified ELO 1674 ± 16, rank #313 of 1776 rated models, from 37 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| SnakeBench | 30.8 | TrueSkill Rating | 88.2 |
| SpeechMap Compliance | 87.7 | % Requests Completed | 86.1 |
| AA-LCR | 64.3 | Score (self-reported) | 82.3 |
| Visual-Language Understanding | 49.15 | Score (self-reported) | 79.2 |
| Wolfram LLM Benchmarking Project | 51.9 | Correct Functionality (%) | 74.8 |
| Chatbot Arena (Text) | 1404 | Elo | 67.6 |
| Chatbot Arena (Vision) | 1225 | Arena Score | 65.7 |
| CritPt | 0.3 | Accuracy (self-reported) | 65 |
| VPCT | 46.2 | Accuracy (%) | 64.1 |
| Vals AI TaxEval v2 | 72.73 | Accuracy (%) | 62.1 |
| Vals AI MedQA | 91.43 | Accuracy (%) | 60.6 |
| Vals AI GPQA | 81.57 | Accuracy (%) | 60.2 |
Interactive version: theaggregate.ai/model?slug=gemini-2-5-flash-preview-09-2025 · How the rankings work · Data refreshed daily, snapshot 2026-07-22.