Model comparison
Llama 2-70B vs Mixtral 8x7B
Mixtral 8x7B is the stronger model overall, scoring 27.1 to 24.4 on the Noometry Index.
Last verified . 30 shared benchmarks.
Meta
Mistral AISummary
- They share 30 benchmarks with published results for both. Llama 2-70B scores higher in 1 category and Mixtral 8x7B in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Mixtral 8x7B leads 18.8 to 8.1.
- The biggest single-benchmark swing is DTBench: 41.6% for Llama 2-70B and 49.6% for Mixtral 8x7B.
Side by side
| Llama 2-70B | Mixtral 8x7B | |
|---|---|---|
| Provider | Meta | Mistral AI |
| Noometry Index | 24.4 | 27.1 |
| Released | 2023-07-18 | 2023-12-11 |
| Weights | Open | Open |
| Context window | — | 32K |
| Max output | — | 32K |
| Input $ / M tokens | — | $0.70 |
| Output $ / M tokens | — | $0.70 |
| Results tracked | 35 | 38 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Mixtral 8x7B leads
Llama 2-70B: 31.4 (#286), Mixtral 8x7B: 32.8 (#269)
| Benchmark | Llama 2-70B | Mixtral 8x7B |
|---|---|---|
| LMArena Coding | 1079 | 1126 |
| HumanEval+ | — | 39.6% |
| MBPP+ | — | 49.7% |
Reasoning Mixtral 8x7B leads
Llama 2-70B: 14.4 (#325), Mixtral 8x7B: 18.2 (#285)
| Benchmark | Llama 2-70B | Mixtral 8x7B |
|---|---|---|
| LMArena Hard Prompts | 1073 | 1115 |
| DTBench | 41.6% | 49.6% |
| Epoch Capabilities Index | 113.79 | 118.47 |
| ForecastBench | 51.4 | 56.3 |
| HellaSwag | 85.3% | 86.7% |
| PIQA | 82.8% | 83.6% |
| WinoGrande | 80.2% | 77.2% |
| Adversarial NLI | — | 55.2% |
| BIG-Bench Hard | 64.9% | — |
| CommonsenseQA 2.0 | 50% | — |
| LAMBADA | 78.9% | — |
Math Mixtral 8x7B leads
Llama 2-70B: 8.1 (#326), Mixtral 8x7B: 18.8 (#289)
| Benchmark | Llama 2-70B | Mixtral 8x7B |
|---|---|---|
| LMArena Math | 1091 | 1147 |
| MATH Level 5 | 3.3% | 10% |
| GSM8K | 69.6% | 74.4% |
| OTIS Mock AIME 2024-2025 | 0% | — |
| Omni-MATH | — | 10.5% |
Knowledge Mixtral 8x7B leads
Llama 2-70B: 7.4 (#310), Mixtral 8x7B: 11.0 (#301)
| Benchmark | Llama 2-70B | Mixtral 8x7B |
|---|---|---|
| GPQA Diamond | 26.3% | 30.6% |
| LMArena Expert | 1039 | 1088 |
| ARC (AI2) Challenge | 78.3% | 87.3% |
| MMLU | 69.9% | 70.6% |
| OpenBookQA | 60.2% | 85.8% |
| TriviaQA | 87.6% | 82.2% |
| MMLU-Pro | — | 33.5% |
| GPQA (HELM) | — | 29.6% |
| BoolQ | 88.6% | — |
Multilingual Mixtral 8x7B leads
Llama 2-70B: 27.7 (#274), Mixtral 8x7B: 29.6 (#266)
| Benchmark | Llama 2-70B | Mixtral 8x7B |
|---|---|---|
| LMArena Non-English | 1045 | 1077 |
| LMArena Chinese | 995 | 1055 |
| LMArena French | 1090 | 1166 |
| LMArena German | 1041 | 1114 |
| LMArena Japanese | 927 | 931 |
| LMArena Korean | 964 | 968 |
| LMArena Russian | 1083 | 1090 |
| LMArena Spanish | 1143 | 1111 |
Instruction Following Llama 2-70B leads
Llama 2-70B: 54.9 (#278), Mixtral 8x7B: 51.0 (#297)
| Benchmark | Llama 2-70B | Mixtral 8x7B |
|---|---|---|
| LMArena Instruction Following | 1071 | 1109 |
| IFEval | — | 57.5% |
Long Context Mixtral 8x7B leads
Llama 2-70B: 32.3 (#270), Mixtral 8x7B: 33.4 (#260)
| Benchmark | Llama 2-70B | Mixtral 8x7B |
|---|---|---|
| LMArena Longer Query | 1062 | 1103 |
Writing & Preference Mixtral 8x7B leads
Llama 2-70B: 32.3 (#279), Mixtral 8x7B: 34.2 (#270)
| Benchmark | Llama 2-70B | Mixtral 8x7B |
|---|---|---|
| LMArena Text | 1115 | 1132 |
| LMArena Creative Writing | 1075 | 1109 |
| LMArena Multi-Turn | 1088 | 1115 |
| WildBench | — | 67.3% |
Frequently asked questions
Is Llama 2-70B better than Mixtral 8x7B?
Mixtral 8x7B is the stronger model overall, scoring 27.1 to 24.4 on the Noometry Index.
Is Llama 2-70B or Mixtral 8x7B better for coding?
Mixtral 8x7B scores higher on coding benchmarks: 32.8 versus 31.4 in the Noometry coding category.
How many benchmarks do Llama 2-70B and Mixtral 8x7B share?
30 benchmarks have published results for both models. Llama 2-70B has 35 scored results on Noometry and Mixtral 8x7B has 38.