Rankings / Mistral AI
Mistral Small 2503
released 2025-03-17
BenchAtlas Index
as of 2026-07-14
base configuration
rank #390
6 families · 3 categories · medium
Benchmark evidence
13 results
external indices
| ECI | 127.8 points | independent | Epoch AI Benchmarking Hub |
knowledge science
| GPQA Diamond | 47.5 % | independent | Epoch AI Benchmarking Hub 2025-03-18 | |
| GPQA Diamond implementation=vals-ai | 44.2 % | independent | Vals AI | |
| MMLU-Pro implementation=vals-ai | 66.0 % | independent | Vals AI |
multimodal
| MMMU implementation=vals-ai | 60.1 % | independent | Vals AI |
professional
| CorpFin | 44.2 % | independent | Vals AI | |
| LegalBench | 69.2 % | independent | Vals AI | |
| MedQA | 69.1 % | independent | Vals AI | |
| TaxEval | 58.3 % | independent | Vals AI |
reasoning math
| AIME implementation=vals-ai | 3.5 % | independent | Vals AI | |
| MATH-500 implementation=vals-ai | 68.4 % | independent | Vals AI | |
| MATH Level 5 | 46.8 % | independent | Epoch AI Benchmarking Hub 2025-03-18 | |
| OTIS Mock AIME 2024–2025 | 5.8 % | independent | Epoch AI Benchmarking Hub 2025-03-18 |
