Model comparison
Llama 3-8B vs Mistral 7B
Llama 3-8B is the stronger model overall, scoring 25.5 to 23.0 on the Noometry Index.
Last verified . 32 shared benchmarks.
Meta
Mistral AISummary
- They share 32 benchmarks with published results for both. Llama 3-8B scores higher in 8 categories and Mistral 7B in 0 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Llama 3-8B leads 37.5 to 30.7.
- The biggest single-benchmark swing is BigCodeBench Instruct: 31.9% for Llama 3-8B and 19.5% for Mistral 7B.
Side by side
| Llama 3-8B | Mistral 7B | |
|---|---|---|
| Provider | Meta | Mistral AI |
| Noometry Index | 25.5 | 23.0 |
| Released | 2024-04-18 | 2023-09-27 |
| Weights | Open | Open |
| Context window | — | 8K |
| Max output | — | 8K |
| Input $ / M tokens | — | $0.25 |
| Output $ / M tokens | — | $0.25 |
| Results tracked | 34 | 37 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Llama 3-8B leads
Llama 3-8B: 31.0 (#289), Mistral 7B: 26.4 (#326)
| Benchmark | Llama 3-8B | Mistral 7B |
|---|---|---|
| BigCodeBench Instruct | 31.9% | 19.5% |
| LMArena Coding | 1152 | 1082 |
| BigCodeBench Complete | 36.9% | 27.3% |
| HumanEval+ | 56.7% | 36% |
| MBPP+ | 54.8% | 42.1% |
Reasoning Llama 3-8B leads
Llama 3-8B: 14.3 (#326), Mistral 7B: 13.1 (#336)
| Benchmark | Llama 3-8B | Mistral 7B |
|---|---|---|
| Chess Puzzles | 0% | 0% |
| LMArena Hard Prompts | 1133 | 1067 |
| DTBench | 43.9% | 42.5% |
| Adversarial NLI | 57.3% | 47.1% |
| Epoch Capabilities Index | 116.45 | 112.21 |
| WinoGrande | 75.7% | 75.3% |
| BIG-Bench Hard | — | 56.1% |
| ForecastBench | 58.6 | — |
| HellaSwag | — | 81% |
| PIQA | — | 83% |
Math Too close to call
Llama 3-8B: 8.8 (#323), Mistral 7B: 8.1 (#325)
| Benchmark | Llama 3-8B | Mistral 7B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.9% | 0.3% |
| LMArena Math | 1151 | 1085 |
| MATH Level 5 | 6.1% | 3.7% |
| GSM8K | — | 54.4% |
Knowledge Too close to call
Llama 3-8B: 7.8 (#308), Mistral 7B: 7.4 (#311)
| Benchmark | Llama 3-8B | Mistral 7B |
|---|---|---|
| GPQA Diamond | 26.1% | 15.2% |
| LMArena Expert | 1113 | 1036 |
| ARC (AI2) Challenge | 82.8% | 78.6% |
| MMLU | 68.8% | 62.5% |
| OpenBookQA | 82.6% | 79.8% |
| TriviaQA | 67.7% | 75.2% |
| BoolQ | — | 87.4% |
Multilingual Llama 3-8B leads
Llama 3-8B: 30.8 (#261), Mistral 7B: 25.8 (#283)
| Benchmark | Llama 3-8B | Mistral 7B |
|---|---|---|
| LMArena Non-English | 1098 | 1012 |
| LMArena Chinese | 1076 | 1009 |
| LMArena French | 1159 | 1037 |
| LMArena German | 1104 | 987 |
| LMArena Japanese | 967 | 878 |
| LMArena Russian | 1109 | 1018 |
| LMArena Spanish | 1173 | 1026 |
| LMArena Korean | 1004 | — |
Instruction Following Llama 3-8B leads
Llama 3-8B: 58.4 (#260), Mistral 7B: 54.2 (#280)
| Benchmark | Llama 3-8B | Mistral 7B |
|---|---|---|
| LMArena Instruction Following | 1127 | 1060 |
Long Context Llama 3-8B leads
Llama 3-8B: 34.2 (#251), Mistral 7B: 32.2 (#271)
| Benchmark | Llama 3-8B | Mistral 7B |
|---|---|---|
| LMArena Longer Query | 1128 | 1060 |
Writing & Preference Llama 3-8B leads
Llama 3-8B: 37.5 (#256), Mistral 7B: 30.7 (#286)
| Benchmark | Llama 3-8B | Mistral 7B |
|---|---|---|
| LMArena Text | 1166 | 1090 |
| LMArena Creative Writing | 1150 | 1068 |
| LMArena Multi-Turn | 1152 | 1062 |
Frequently asked questions
Is Llama 3-8B better than Mistral 7B?
Llama 3-8B is the stronger model overall, scoring 25.5 to 23.0 on the Noometry Index.
Is Llama 3-8B or Mistral 7B better for coding?
Llama 3-8B scores higher on coding benchmarks: 31.0 versus 26.4 in the Noometry coding category.
How many benchmarks do Llama 3-8B and Mistral 7B share?
32 benchmarks have published results for both models. Llama 3-8B has 34 scored results on Noometry and Mistral 7B has 37.