Gemini 2.5 Flash May 2025
Gemini 2.5 Flash.
17
Benchmarks with a score
0
Ranked first, out of five or more models
0
In the top three
Google DeepMind May 2025 Gemini 2.5 Flash
May 2025 snapshot Closed weights
141.6
#115 of 256
Epoch Capabilities Index
- Humanity's Last Exam 11% #28 /47
- VISTA 49.1% #13 /54
- Misguided Attention 63.5% #5 /45
Where it ranks
Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.
Knowledge and language
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Fiction.LiveBench | 77.8% | #13 of 57 | Fiction.live leaderboard, via Epoch AI | |
| Humanity's Last Exam | 11% | #28 of 47 | Official leaderboard |
Math and science
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Misguided Attention | 63.5% | #5 of 45 | Official leaderboard | |
| IneqMath | 27.5% | #5 of 34 | Official leaderboard | |
| OTIS Mock AIME 2024-2025 | 70.8% | #109 of 212 | Epoch AI run | |
| ARC-AGI-2 | 2.54% | #77 of 98 | Official leaderboard | |
| ARC-AGI | 33.3% | #76 of 95 | Official leaderboard |
Coding
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Aider Polyglot | 55.1% | #18 of 46 | Official leaderboard | |
| LiveCodeBench | 82.9% | #11 of 21 | Official leaderboard | |
| LiveCodeBench v5 | 62.8% | #12 of 21 | Official leaderboard | |
| LiveCodeBench v6 | 61.9% | #12 of 21 | Official leaderboard |
Professional work
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| xbench-Recruitment | 60.6 | #4 of 8 | Official leaderboard | |
| xbench-Marketing | 45.3 | #5 of 7 | Official leaderboard | |
| TaxCalcBench | 26% | #15 of 17 | Official leaderboard |
Google DeepMind's models over time
Google DeepMind's models in color, against every other model on the Epoch Capabilities Index, by release date.
Source: Epoch Capabilities Index, Epoch AI (CC BY 4.0)