Mistral 7B v0.3
Mistral 7B v0.3.
Mistral AI Released 27 Sep 2023 24
Benchmarks with a score
0
Ranked first, out of five or more models
0
In the top three
Mistral AI Sep 2023 Mistral 7B v0.3
Open weights
109.0
#227 of 271
Epoch Capabilities Index
- GPQA Diamond 15.2% #216 /219
- KOR-Bench 24.2% #31 /33
- AgentIF 46.8% #12 /12
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 | |
|---|---|---|---|---|
| Vectara Hallucination Leaderboard (HHEM) | 9.50% | #21 of 148 | Official leaderboard | |
| RepoQA | 62% | #14 of 22 | Official leaderboard | |
| MMLU | 59.9% | #83 of 127 | Epoch AI | |
| AlpacaEval 2.0 | 20.6% | #28 of 40 | Official leaderboard | |
| EQ-Bench v2 | 63.1 | #62 of 85 | Official leaderboard | |
| MMLU-CF | 50.7% | #31 of 37 | Official leaderboard | |
| ThaiExam | 32% | #21 of 24 | Official leaderboard | |
| ToRR | 32.4 | #13 of 14 | Official leaderboard | |
| SuperGPQA | 17.8% | #57 of 58 | Official leaderboard | |
| AgentIF | 46.8% | #12 of 12 | Official leaderboard |
Math and science
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| KOR-Bench | 24.2% | #31 of 33 | Official leaderboard | |
| GPQA Diamond | 15.2% | #216 of 219 | Epoch AI | |
| DTBench | 42.5% | #175 of 177 | Epoch AI | |
| MATH Level 5 | 3.68% | #93 of 94 | Epoch AI | |
| OTIS Mock AIME 2024-2025 | 0.28% | #200 of 202 | Epoch AI |
Coding
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| BigCodeBench-Hard | 5.40% | #90 of 96 | Official leaderboard | |
| BigCodeBench | 23.4% | #65 of 69 | Official leaderboard | |
| MHPP | 9.00% | #23 of 23 | Official leaderboard |
Safety and security
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| AIR-Bench 2024 | 32.5% | #79 of 82 | Official leaderboard | |
| Anthropic Red Team | 90.1% | #73 of 75 | Official leaderboard | |
| HELM Safety | 73.0 | #73 of 75 | Official leaderboard | |
| SimpleSafetyTests | 81% | #73 of 75 | Official leaderboard |
Other
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| HELM Lite | 19.6 | #68 of 71 | Official leaderboard | |
| HELM Capabilities | 37.6 | #60 of 60 | Official leaderboard |
Mistral AI's models over time
Mistral AI's models in color, against every other model on the Epoch Capabilities Index, by release date.