Model comparison
Llama 3-8B vs Mistral
Mistral is the stronger model overall, scoring 29.4 to 25.3 on the Noometry Index.
Last verified . 17 shared benchmarks.
Meta
Mistral AISummary
- They share 17 benchmarks with published results for both. Llama 3-8B scores higher in 2 categories and Mistral in 6 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in math, where Mistral leads 20.5 to 8.3.
- Llama 3-8B has downloadable open weights; the other is API-only.
Side by side
| Llama 3-8B | Mistral | |
|---|---|---|
| Provider | Meta | Mistral AI |
| Noometry Index | 25.3 | 29.4 |
| Released | 2024-04-18 | — |
| Weights | Open | Proprietary |
| Context window | 8K | — |
| Max output | 8K | — |
| Input $ / M tokens | $0.14 | — |
| Output $ / M tokens | $0.14 | — |
| Results tracked | 34 | 22 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Mistral leads
Llama 3-8B: 30.5 (#285), Mistral: 33.2 (#249)
| Benchmark | Llama 3-8B | Mistral |
|---|---|---|
| LMArena Coding | 1152 | 1162 |
| BigCodeBench Instruct | 31.9% | — |
| BigCodeBench Complete | 36.9% | — |
| HumanEval+ | 56.7% | — |
| MBPP+ | 54.8% | — |
Reasoning Mistral leads
Llama 3-8B: 13.8 (#321), Mistral: 21.3 (#199)
| Benchmark | Llama 3-8B | Mistral |
|---|---|---|
| LMArena Hard Prompts | 1133 | 1149 |
| Chess Puzzles | 0% | — |
| DTBench | 43.9% | — |
| Adversarial NLI | 57.3% | — |
| Epoch Capabilities Index | 116.33 | — |
| ForecastBench | 58.6 | — |
| WinoGrande | 75.7% | — |
Math Mistral leads
Llama 3-8B: 8.3 (#317), Mistral: 20.5 (#275)
| Benchmark | Llama 3-8B | Mistral |
|---|---|---|
| LMArena Math | 1151 | 1180 |
| OTIS Mock AIME 2024-2025 | 1.9% | — |
| Omni-MATH | — | 7.2% |
| MATH Level 5 | 6.1% | — |
Knowledge Mistral leads
Llama 3-8B: 7.8 (#303), Mistral: 16.3 (#283)
| Benchmark | Llama 3-8B | Mistral |
|---|---|---|
| LMArena Expert | 1113 | 1125 |
| GPQA Diamond | 26.1% | — |
| MMLU-Pro | — | 27.7% |
| GPQA (HELM) | — | 30.3% |
| ARC (AI2) Challenge | 82.8% | — |
| MMLU | 68.8% | — |
| OpenBookQA | 82.6% | — |
| TriviaQA | 67.7% | — |
Multilingual Mistral leads
Llama 3-8B: 31.0 (#256), Mistral: 33.0 (#249)
| Benchmark | Llama 3-8B | Mistral |
|---|---|---|
| LMArena Non-English | 1098 | 1129 |
| LMArena Chinese | 1076 | 1109 |
| LMArena French | 1158 | 1179 |
| LMArena German | 1104 | 1155 |
| LMArena Japanese | 967 | 1012 |
| LMArena Korean | 1003 | 1032 |
| LMArena Russian | 1109 | 1168 |
| LMArena Spanish | 1172 | 1143 |
Instruction Following Llama 3-8B leads
Llama 3-8B: 58.7 (#255), Mistral: 52.9 (#283)
| Benchmark | Llama 3-8B | Mistral |
|---|---|---|
| LMArena Instruction Following | 1127 | 1152 |
| IFEval | — | 56.8% |
Long Context Too close to call
Llama 3-8B: 34.2 (#246), Mistral: 35.0 (#240)
| Benchmark | Llama 3-8B | Mistral |
|---|---|---|
| LMArena Longer Query | 1128 | 1154 |
Writing & Preference Too close to call
Llama 3-8B: 37.7 (#251), Mistral: 37.2 (#255)
| Benchmark | Llama 3-8B | Mistral |
|---|---|---|
| LMArena Text | 1166 | 1165 |
| LMArena Creative Writing | 1150 | 1158 |
| LMArena Multi-Turn | 1152 | 1147 |
| WildBench | — | 66% |
Frequently asked questions
Is Llama 3-8B better than Mistral?
Mistral is the stronger model overall, scoring 29.4 to 25.3 on the Noometry Index.
Is Llama 3-8B or Mistral better for coding?
Mistral scores higher on coding benchmarks: 33.2 versus 30.5 in the Noometry coding category.
How many benchmarks do Llama 3-8B and Mistral share?
17 benchmarks have published results for both models. Llama 3-8B has 34 scored results on Noometry and Mistral has 22.