Model comparison
Mixtral 8x7B vs Qwen1.5-32B
Qwen1.5-32B is the stronger model overall, scoring 30.5 to 27.1 on the Noometry Index.
Last verified . 19 shared benchmarks.
Mistral AI
Alibaba (Qwen)Summary
- They share 19 benchmarks with published results for both. Mixtral 8x7B scores higher in 2 categories and Qwen1.5-32B in 6 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen1.5-32B leads 33.0 to 18.8.
Side by side
| Mixtral 8x7B | Qwen1.5-32B | |
|---|---|---|
| Provider | Mistral AI | Alibaba (Qwen) |
| Noometry Index | 27.1 | 30.5 |
| Released | 2023-12-11 | 2024-02-04 |
| Weights | Open | Open |
| Context window | 32K | — |
| Max output | 32K | — |
| Input $ / M tokens | $0.70 | — |
| Output $ / M tokens | $0.70 | — |
| Results tracked | 38 | 21 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Mixtral 8x7B leads
Mixtral 8x7B: 32.8 (#269), Qwen1.5-32B: 31.7 (#282)
| Benchmark | Mixtral 8x7B | Qwen1.5-32B |
|---|---|---|
| LMArena Coding | 1126 | 1155 |
| BigCodeBench Instruct | — | 32.3% |
| BigCodeBench Complete | — | 42% |
| HumanEval+ | 39.6% | — |
| MBPP+ | 49.7% | — |
Reasoning Qwen1.5-32B leads
Mixtral 8x7B: 18.2 (#285), Qwen1.5-32B: 21.8 (#212)
| Benchmark | Mixtral 8x7B | Qwen1.5-32B |
|---|---|---|
| LMArena Hard Prompts | 1115 | 1130 |
| DTBench | 49.6% | — |
| Adversarial NLI | 55.2% | — |
| Epoch Capabilities Index | 118.47 | — |
| ForecastBench | 56.3 | — |
| HellaSwag | 86.7% | — |
| PIQA | 83.6% | — |
| WinoGrande | 77.2% | — |
Math Qwen1.5-32B leads
Mixtral 8x7B: 18.8 (#289), Qwen1.5-32B: 33.0 (#207)
| Benchmark | Mixtral 8x7B | Qwen1.5-32B |
|---|---|---|
| LMArena Math | 1147 | 1155 |
| Omni-MATH | 10.5% | — |
| MATH Level 5 | 10% | — |
| GSM8K | 74.4% | — |
Knowledge Qwen1.5-32B leads
Mixtral 8x7B: 11.0 (#301), Qwen1.5-32B: 13.5 (#296)
| Benchmark | Mixtral 8x7B | Qwen1.5-32B |
|---|---|---|
| GPQA Diamond | 30.6% | 30.7% |
| LMArena Expert | 1088 | 1126 |
| MMLU | 70.6% | 74.4% |
| MMLU-Pro | 33.5% | — |
| GPQA (HELM) | 29.6% | — |
| ARC (AI2) Challenge | 87.3% | — |
| OpenBookQA | 85.8% | — |
| TriviaQA | 82.2% | — |
Multilingual Qwen1.5-32B leads
Mixtral 8x7B: 29.6 (#266), Qwen1.5-32B: 31.4 (#259)
| Benchmark | Mixtral 8x7B | Qwen1.5-32B |
|---|---|---|
| LMArena Non-English | 1077 | 1106 |
| LMArena Chinese | 1055 | 1177 |
| LMArena French | 1166 | 1101 |
| LMArena German | 1114 | 1058 |
| LMArena Japanese | 931 | 1027 |
| LMArena Korean | 968 | 1008 |
| LMArena Russian | 1090 | 1073 |
| LMArena Spanish | 1111 | 1089 |
Instruction Following Qwen1.5-32B leads
Mixtral 8x7B: 51.0 (#297), Qwen1.5-32B: 57.7 (#265)
| Benchmark | Mixtral 8x7B | Qwen1.5-32B |
|---|---|---|
| LMArena Instruction Following | 1109 | 1116 |
| IFEval | 57.5% | — |
Long Context Qwen1.5-32B leads
Mixtral 8x7B: 33.4 (#260), Qwen1.5-32B: 34.7 (#246)
| Benchmark | Mixtral 8x7B | Qwen1.5-32B |
|---|---|---|
| LMArena Longer Query | 1103 | 1146 |
Writing & Preference Too close to call
Mixtral 8x7B: 34.2 (#270), Qwen1.5-32B: 34.2 (#271)
| Benchmark | Mixtral 8x7B | Qwen1.5-32B |
|---|---|---|
| LMArena Text | 1132 | 1137 |
| LMArena Creative Writing | 1109 | 1083 |
| LMArena Multi-Turn | 1115 | 1140 |
| WildBench | 67.3% | — |
Frequently asked questions
Is Mixtral 8x7B better than Qwen1.5-32B?
Qwen1.5-32B is the stronger model overall, scoring 30.5 to 27.1 on the Noometry Index.
Is Mixtral 8x7B or Qwen1.5-32B better for coding?
Mixtral 8x7B scores higher on coding benchmarks: 32.8 versus 31.7 in the Noometry coding category.
How many benchmarks do Mixtral 8x7B and Qwen1.5-32B share?
19 benchmarks have published results for both models. Mixtral 8x7B has 38 scored results on Noometry and Qwen1.5-32B has 21.