Model comparison
Llama 2-70B vs Mistral 7B
Llama 2-70B is the stronger model overall, scoring 24.4 to 23.0 on the Noometry Index.
Last verified . 31 shared benchmarks.
Meta
Mistral AISummary
- They share 31 benchmarks with published results for both. Llama 2-70B scores higher in 7 categories and Mistral 7B in 1 category; 4 gaps are clear of the uncertainty.
- The widest gap is in coding, where Llama 2-70B leads 31.4 to 26.4.
- The biggest single-benchmark swing is GPQA Diamond: 26.3% for Llama 2-70B and 15.2% for Mistral 7B.
Side by side
| Llama 2-70B | Mistral 7B | |
|---|---|---|
| Provider | Meta | Mistral AI |
| Noometry Index | 24.4 | 23.0 |
| Released | 2023-07-18 | 2023-09-27 |
| Weights | Open | Open |
| Context window | — | 8K |
| Max output | — | 8K |
| Input $ / M tokens | — | $0.25 |
| Output $ / M tokens | — | $0.25 |
| Results tracked | 35 | 37 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Llama 2-70B leads
Llama 2-70B: 31.4 (#286), Mistral 7B: 26.4 (#326)
| Benchmark | Llama 2-70B | Mistral 7B |
|---|---|---|
| LMArena Coding | 1079 | 1082 |
| BigCodeBench Instruct | — | 19.5% |
| BigCodeBench Complete | — | 27.3% |
| HumanEval+ | — | 36% |
| MBPP+ | — | 42.1% |
Reasoning Llama 2-70B leads
Llama 2-70B: 14.4 (#325), Mistral 7B: 13.1 (#336)
| Benchmark | Llama 2-70B | Mistral 7B |
|---|---|---|
| LMArena Hard Prompts | 1073 | 1067 |
| DTBench | 41.6% | 42.5% |
| BIG-Bench Hard | 64.9% | 56.1% |
| Epoch Capabilities Index | 113.79 | 112.21 |
| HellaSwag | 85.3% | 81% |
| PIQA | 82.8% | 83% |
| WinoGrande | 80.2% | 75.3% |
| Chess Puzzles | — | 0% |
| Adversarial NLI | — | 47.1% |
| CommonsenseQA 2.0 | 50% | — |
| ForecastBench | 51.4 | — |
| LAMBADA | 78.9% | — |
Math Too close to call
Llama 2-70B: 8.1 (#326), Mistral 7B: 8.1 (#325)
| Benchmark | Llama 2-70B | Mistral 7B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 0% | 0.3% |
| LMArena Math | 1091 | 1085 |
| MATH Level 5 | 3.3% | 3.7% |
| GSM8K | 69.6% | 54.4% |
Knowledge Too close to call
Llama 2-70B: 7.4 (#310), Mistral 7B: 7.4 (#311)
| Benchmark | Llama 2-70B | Mistral 7B |
|---|---|---|
| GPQA Diamond | 26.3% | 15.2% |
| LMArena Expert | 1039 | 1036 |
| ARC (AI2) Challenge | 78.3% | 78.6% |
| BoolQ | 88.6% | 87.4% |
| MMLU | 69.9% | 62.5% |
| OpenBookQA | 60.2% | 79.8% |
| TriviaQA | 87.6% | 75.2% |
Multilingual Llama 2-70B leads
Llama 2-70B: 27.7 (#274), Mistral 7B: 25.8 (#283)
| Benchmark | Llama 2-70B | Mistral 7B |
|---|---|---|
| LMArena Non-English | 1045 | 1012 |
| LMArena Chinese | 995 | 1009 |
| LMArena French | 1090 | 1037 |
| LMArena German | 1041 | 987 |
| LMArena Japanese | 927 | 878 |
| LMArena Russian | 1083 | 1018 |
| LMArena Spanish | 1143 | 1026 |
| LMArena Korean | 964 | — |
Instruction Following Too close to call
Llama 2-70B: 54.9 (#278), Mistral 7B: 54.2 (#280)
| Benchmark | Llama 2-70B | Mistral 7B |
|---|---|---|
| LMArena Instruction Following | 1071 | 1060 |
Long Context Too close to call
Llama 2-70B: 32.3 (#270), Mistral 7B: 32.2 (#271)
| Benchmark | Llama 2-70B | Mistral 7B |
|---|---|---|
| LMArena Longer Query | 1062 | 1060 |
Writing & Preference Llama 2-70B leads
Llama 2-70B: 32.3 (#279), Mistral 7B: 30.7 (#286)
| Benchmark | Llama 2-70B | Mistral 7B |
|---|---|---|
| LMArena Text | 1115 | 1090 |
| LMArena Creative Writing | 1075 | 1068 |
| LMArena Multi-Turn | 1088 | 1062 |
Frequently asked questions
Is Llama 2-70B better than Mistral 7B?
Llama 2-70B is the stronger model overall, scoring 24.4 to 23.0 on the Noometry Index.
Is Llama 2-70B or Mistral 7B better for coding?
Llama 2-70B scores higher on coding benchmarks: 31.4 versus 26.4 in the Noometry coding category.
How many benchmarks do Llama 2-70B and Mistral 7B share?
31 benchmarks have published results for both models. Llama 2-70B has 35 scored results on Noometry and Mistral 7B has 37.