Rankings / Mistral AI
Mistral Medium 3
released 2025-05-07
BenchAtlas Index
as of 2026-07-14
base configuration
rank #399
11 families · 6 categories · high
Benchmark evidence
16 results
agentic coding
| τ²-Bench | 24.3 % | independent | Artificial Analysis |
coding
| LiveCodeBench v6 implementation=artificial-analysis | 40.0 % | independent | Artificial Analysis | |
| SciCode | 33.1 % | independent | Artificial Analysis |
external indices
| Intelligence Index v4.1 | 12.5 points | independent | Artificial Analysis | |
| AA Math Index | 30.3 points | independent | Artificial Analysis |
factuality
| MASK | 42.6 % 39.3–45.9 | independent | Scale Labs |
knowledge science
| GPQA Diamond implementation=artificial-analysis | 57.8 % | independent | Artificial Analysis | |
| Humanity's Last Exam contamination=Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions. · implementation=scale | 4.5 % 3.7–5.3 | independent | Scale Labs | |
| Humanity's Last Exam implementation=artificial-analysis | 4.3 % | independent | Artificial Analysis | |
| MMLU-Pro implementation=artificial-analysis | 76.0 % | independent | Artificial Analysis |
long context instruction
| AA-LCR | 28.0 % | independent | Artificial Analysis | |
| IFBench | 39.3 % | independent | Artificial Analysis |
multimodal
| VISTA | 34.6 % 33.5–35.7 | independent | Scale Labs |
reasoning math
| AIME implementation=artificial-analysis | 44.0 % | independent | Artificial Analysis | |
| AIME year=2025 · implementation=artificial-analysis | 30.3 % | independent | Artificial Analysis | |
| MATH-500 implementation=artificial-analysis | 90.7 % | independent | Artificial Analysis |
Agent + model results
systems, not bare-model scores
| agent + model Artificial Analysis harness + Mistral Medium 3 | Terminal-Bench Hard | 3.8 % | independent | Artificial Analysis |
These scores measure the whole agent system (scaffold, tools, budgets) — they are never merged into the bare model’s numbers.
