Model comparison
Llama 3.1-405B vs Mixtral 8x7B
Llama 3.1-405B is the stronger model overall, scoring 30.7 to 27.1 on the Noometry Index.
Last verified . 33 shared benchmarks.
Meta
Mistral AISummary
- They share 33 benchmarks with published results for both. Llama 3.1-405B scores higher in 6 categories and Mixtral 8x7B in 2 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Llama 3.1-405B leads 30.4 to 11.0.
- The biggest single-benchmark swing is MATH Level 5: 49.8% for Llama 3.1-405B and 10% for Mixtral 8x7B.
Side by side
| Llama 3.1-405B | Mixtral 8x7B | |
|---|---|---|
| Provider | Meta | Mistral AI |
| Noometry Index | 30.7 | 27.1 |
| Released | 2024-07-23 | 2023-12-11 |
| Weights | Open | Open |
| Context window | — | 32K |
| Max output | — | 32K |
| Input $ / M tokens | — | $0.70 |
| Output $ / M tokens | — | $0.70 |
| Results tracked | 42 | 38 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Llama 3.1-405B: 33.1 (#262), Mixtral 8x7B: 32.8 (#269)
| Benchmark | Llama 3.1-405B | Mixtral 8x7B |
|---|---|---|
| LMArena Coding | 1291 | 1126 |
| WeirdML | 21.4% | — |
| HumanEval+ | — | 39.6% |
| MBPP+ | — | 49.7% |
Agentic & Tool Use Not comparable
Llama 3.1-405B: 21.0 (#140), Mixtral 8x7B: —
| Benchmark | Llama 3.1-405B | Mixtral 8x7B |
|---|---|---|
| TheAgentCompany | 7.4% | — |
| Cybench | 7.5% | — |
Reasoning Mixtral 8x7B leads
Llama 3.1-405B: 16.8 (#300), Mixtral 8x7B: 18.2 (#285)
| Benchmark | Llama 3.1-405B | Mixtral 8x7B |
|---|---|---|
| LMArena Hard Prompts | 1269 | 1115 |
| DTBench | 61.4% | 49.6% |
| Epoch Capabilities Index | 128.75 | 118.47 |
| ForecastBench | 59.9 | 56.3 |
| HellaSwag | 89.2% | 86.7% |
| PIQA | 85.9% | 83.6% |
| WinoGrande | 89.2% | 77.2% |
| SimpleBench | 23% | — |
| Kagi LLM Benchmark | 45% | — |
| Adversarial NLI | — | 55.2% |
| BIG-Bench Hard | 82.9% | — |
Math Too close to call
Llama 3.1-405B: 18.4 (#290), Mixtral 8x7B: 18.8 (#289)
| Benchmark | Llama 3.1-405B | Mixtral 8x7B |
|---|---|---|
| Omni-MATH | 24.9% | 10.5% |
| LMArena Math | 1281 | 1147 |
| MATH Level 5 | 49.8% | 10% |
| OTIS Mock AIME 2024-2025 | 9.7% | — |
| GSM8K | — | 74.4% |
Knowledge Llama 3.1-405B leads
Llama 3.1-405B: 30.4 (#227), Mixtral 8x7B: 11.0 (#301)
| Benchmark | Llama 3.1-405B | Mixtral 8x7B |
|---|---|---|
| GPQA Diamond | 50.9% | 30.6% |
| MMLU-Pro | 72.3% | 33.5% |
| GPQA (HELM) | 52.2% | 29.6% |
| LMArena Expert | 1243 | 1088 |
| ARC (AI2) Challenge | 95.3% | 87.3% |
| MMLU | 84.5% | 70.6% |
| TriviaQA | 82.7% | 82.2% |
| Confabulations | 17.6% | — |
| OpenBookQA | — | 85.8% |
Multilingual Llama 3.1-405B leads
Llama 3.1-405B: 40.7 (#214), Mixtral 8x7B: 29.6 (#266)
| Benchmark | Llama 3.1-405B | Mixtral 8x7B |
|---|---|---|
| LMArena Non-English | 1248 | 1077 |
| LMArena Chinese | 1242 | 1055 |
| LMArena French | 1279 | 1166 |
| LMArena German | 1252 | 1114 |
| LMArena Japanese | 1208 | 931 |
| LMArena Korean | 1184 | 968 |
| LMArena Russian | 1265 | 1090 |
| LMArena Spanish | 1260 | 1111 |
Instruction Following Llama 3.1-405B leads
Llama 3.1-405B: 65.9 (#214), Mixtral 8x7B: 51.0 (#297)
| Benchmark | Llama 3.1-405B | Mixtral 8x7B |
|---|---|---|
| IFEval | 81.1% | 57.5% |
| LMArena Instruction Following | 1259 | 1109 |
Long Context Llama 3.1-405B leads
Llama 3.1-405B: 38.4 (#197), Mixtral 8x7B: 33.4 (#260)
| Benchmark | Llama 3.1-405B | Mixtral 8x7B |
|---|---|---|
| LMArena Longer Query | 1266 | 1103 |
Writing & Preference Llama 3.1-405B leads
Llama 3.1-405B: 38.9 (#251), Mixtral 8x7B: 34.2 (#270)
| Benchmark | Llama 3.1-405B | Mixtral 8x7B |
|---|---|---|
| LMArena Text | 1284 | 1132 |
| LMArena Creative Writing | 1262 | 1109 |
| WildBench | 78.3% | 67.3% |
| LMArena Multi-Turn | 1297 | 1115 |
| EQ-Bench Creative Writing | 870 | — |
Frequently asked questions
Is Llama 3.1-405B better than Mixtral 8x7B?
Llama 3.1-405B is the stronger model overall, scoring 30.7 to 27.1 on the Noometry Index.
Is Llama 3.1-405B or Mixtral 8x7B better for coding?
They score almost the same on coding (33.1 vs 32.8); test both on your own repository before choosing.
How many benchmarks do Llama 3.1-405B and Mixtral 8x7B share?
33 benchmarks have published results for both models. Llama 3.1-405B has 42 scored results on Noometry and Mixtral 8x7B has 38.