Rankings / Google DeepMind
Gemini 2.5 Pro Exp 03.25
released 2025-03-25
BenchAtlas Index
as of 2026-07-14
base configuration
rank #105
7 families · 4 categories · medium
Benchmark evidence
11 results
external indices
| ECI | 144.8 points | independent | Epoch AI Benchmarking Hub |
knowledge science
| GPQA Diamond | 83.8 % | independent | Epoch AI Benchmarking Hub 2025-03-31 | |
| GPQA Diamond implementation=vals-ai | 80.8 % | independent | Vals AI | |
| MMLU-Pro implementation=vals-ai | 84.1 % | independent | Vals AI |
multimodal
| MMMU implementation=vals-ai | 81.3 % | independent | Vals AI |
professional
| CorpFin | 59.8 % | independent | Vals AI | |
| LegalBench | 84.3 % | independent | Vals AI | |
| MedQA | 93.1 % | independent | Vals AI | |
| TaxEval | 72.9 % | independent | Vals AI |
reasoning math
| AIME implementation=vals-ai | 85.8 % | independent | Vals AI | |
| MATH-500 implementation=vals-ai | 95.2 % | independent | Vals AI |