Model comparison
Llama 3.1-405B vs Mixtral 8x22B
Llama 3.1-405B is the stronger model overall, scoring 30.7 to 27.1 on the Noometry Index.
Last verified . 30 shared benchmarks.
Meta
Mistral AISummary
- They share 30 benchmarks with published results for both. Llama 3.1-405B scores higher in 6 categories and Mixtral 8x22B in 3 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Llama 3.1-405B leads 30.4 to 15.1.
- The biggest single-benchmark swing is MMLU-Pro: 72.3% for Llama 3.1-405B and 46% for Mixtral 8x22B.
Side by side
| Llama 3.1-405B | Mixtral 8x22B | |
|---|---|---|
| Provider | Meta | Mistral AI |
| Noometry Index | 30.7 | 27.1 |
| Released | 2024-07-23 | 2024-04-17 |
| Weights | Open | Open |
| Context window | — | 64K |
| Max output | — | 64K |
| Input $ / M tokens | — | $2 |
| Output $ / M tokens | — | $6 |
| Results tracked | 42 | 34 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Llama 3.1-405B leads
Llama 3.1-405B: 33.1 (#262), Mixtral 8x22B: 24.2 (#329)
| Benchmark | Llama 3.1-405B | Mixtral 8x22B |
|---|---|---|
| WeirdML | 21.4% | 3.2% |
| LMArena Coding | 1291 | 1166 |
| BigCodeBench Instruct | — | 40.6% |
| BigCodeBench Complete | — | 50.2% |
| HumanEval+ | — | 72% |
| MBPP+ | — | 64.3% |
Agentic & Tool Use Mixtral 8x22B leads
Llama 3.1-405B: 21.0 (#140), Mixtral 8x22B: 23.1 (#127)
| Benchmark | Llama 3.1-405B | Mixtral 8x22B |
|---|---|---|
| Cybench | 7.5% | 7.5% |
| TheAgentCompany | 7.4% | — |
Reasoning Mixtral 8x22B leads
Llama 3.1-405B: 16.8 (#300), Mixtral 8x22B: 19.9 (#248)
| Benchmark | Llama 3.1-405B | Mixtral 8x22B |
|---|---|---|
| LMArena Hard Prompts | 1269 | 1150 |
| DTBench | 61.4% | 55.1% |
| Epoch Capabilities Index | 128.75 | 122.03 |
| ForecastBench | 59.9 | 56.3 |
| SimpleBench | 23% | — |
| Kagi LLM Benchmark | 45% | — |
| BIG-Bench Hard | 82.9% | — |
| HellaSwag | 89.2% | — |
| PIQA | 85.9% | — |
| WinoGrande | 89.2% | — |
Math Mixtral 8x22B leads
Llama 3.1-405B: 18.4 (#290), Mixtral 8x22B: 22.9 (#275)
| Benchmark | Llama 3.1-405B | Mixtral 8x22B |
|---|---|---|
| Omni-MATH | 24.9% | 16.3% |
| LMArena Math | 1281 | 1184 |
| MATH Level 5 | 49.8% | 24.2% |
| OTIS Mock AIME 2024-2025 | 9.7% | — |
Knowledge Llama 3.1-405B leads
Llama 3.1-405B: 30.4 (#227), Mixtral 8x22B: 15.1 (#293)
| Benchmark | Llama 3.1-405B | Mixtral 8x22B |
|---|---|---|
| GPQA Diamond | 50.9% | 34.1% |
| MMLU-Pro | 72.3% | 46% |
| GPQA (HELM) | 52.2% | 33.4% |
| LMArena Expert | 1243 | 1113 |
| MMLU | 84.5% | 77.8% |
| Confabulations | 17.6% | — |
| ARC (AI2) Challenge | 95.3% | — |
| TriviaQA | 82.7% | — |
Multilingual Llama 3.1-405B leads
Llama 3.1-405B: 40.7 (#214), Mixtral 8x22B: 32.8 (#255)
| Benchmark | Llama 3.1-405B | Mixtral 8x22B |
|---|---|---|
| LMArena Non-English | 1248 | 1128 |
| LMArena Chinese | 1242 | 1116 |
| LMArena French | 1279 | 1166 |
| LMArena German | 1252 | 1141 |
| LMArena Japanese | 1208 | 1037 |
| LMArena Korean | 1184 | 1057 |
| LMArena Russian | 1265 | 1158 |
| LMArena Spanish | 1260 | 1151 |
Instruction Following Llama 3.1-405B leads
Llama 3.1-405B: 65.9 (#214), Mixtral 8x22B: 57.7 (#266)
| Benchmark | Llama 3.1-405B | Mixtral 8x22B |
|---|---|---|
| IFEval | 81.1% | 72.4% |
| LMArena Instruction Following | 1259 | 1147 |
Long Context Llama 3.1-405B leads
Llama 3.1-405B: 38.4 (#197), Mixtral 8x22B: 34.7 (#247)
| Benchmark | Llama 3.1-405B | Mixtral 8x22B |
|---|---|---|
| LMArena Longer Query | 1266 | 1144 |
Writing & Preference Llama 3.1-405B leads
Llama 3.1-405B: 38.9 (#251), Mixtral 8x22B: 36.9 (#262)
| Benchmark | Llama 3.1-405B | Mixtral 8x22B |
|---|---|---|
| LMArena Text | 1284 | 1162 |
| LMArena Creative Writing | 1262 | 1141 |
| WildBench | 78.3% | 71.1% |
| LMArena Multi-Turn | 1297 | 1130 |
| EQ-Bench Creative Writing | 870 | — |
Frequently asked questions
Is Llama 3.1-405B better than Mixtral 8x22B?
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 8x22B better for coding?
Llama 3.1-405B scores higher on coding benchmarks: 33.1 versus 24.2 in the Noometry coding category.
How many benchmarks do Llama 3.1-405B and Mixtral 8x22B share?
30 benchmarks have published results for both models. Llama 3.1-405B has 42 scored results on Noometry and Mixtral 8x22B has 34.