Model comparison
Mixtral 8x22B vs Qwen2-72B
Qwen2-72B is the stronger model overall, scoring 30.0 to 27.1 on the Noometry Index.
Last verified . 24 shared benchmarks.
Mistral AI
Alibaba (Qwen)Summary
- They share 24 benchmarks with published results for both. Mixtral 8x22B scores higher in 1 category and Qwen2-72B in 8 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen2-72B leads 30.2 to 22.9.
- The biggest single-benchmark swing is MATH Level 5: 24.2% for Mixtral 8x22B and 39.1% for Qwen2-72B.
Side by side
| Mixtral 8x22B | Qwen2-72B | |
|---|---|---|
| Provider | Mistral AI | Alibaba (Qwen) |
| Noometry Index | 27.1 | 30.0 |
| Released | 2024-04-17 | 2024-06-07 |
| Weights | Open | Open |
| Context window | 64K | — |
| Max output | 64K | — |
| Input $ / M tokens | $2 | — |
| Output $ / M tokens | $6 | — |
| Results tracked | 34 | 26 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen2-72B leads
Mixtral 8x22B: 24.2 (#329), Qwen2-72B: 29.1 (#310)
| Benchmark | Mixtral 8x22B | Qwen2-72B |
|---|---|---|
| WeirdML | 3.2% | 11.3% |
| BigCodeBench Instruct | 40.6% | 38.5% |
| LMArena Coding | 1166 | 1196 |
| BigCodeBench Complete | 50.2% | 54% |
| HumanEval+ | 72% | — |
| MBPP+ | 64.3% | — |
Agentic & Tool Use Mixtral 8x22B leads
Mixtral 8x22B: 23.1 (#127), Qwen2-72B: 17.0 (#146)
| Benchmark | Mixtral 8x22B | Qwen2-72B |
|---|---|---|
| TheAgentCompany | — | 1.1% |
| Cybench | 7.5% | — |
| METR Time Horizons | — | 29.9% |
Reasoning Qwen2-72B leads
Mixtral 8x22B: 19.9 (#248), Qwen2-72B: 23.2 (#181)
| Benchmark | Mixtral 8x22B | Qwen2-72B |
|---|---|---|
| LMArena Hard Prompts | 1150 | 1191 |
| Epoch Capabilities Index | 122.03 | 125.28 |
| DTBench | 55.1% | — |
| ForecastBench | 56.3 | — |
Math Qwen2-72B leads
Mixtral 8x22B: 22.9 (#275), Qwen2-72B: 30.2 (#236)
| Benchmark | Mixtral 8x22B | Qwen2-72B |
|---|---|---|
| LMArena Math | 1184 | 1235 |
| MATH Level 5 | 24.2% | 39.1% |
| Omni-MATH | 16.3% | — |
Knowledge Qwen2-72B leads
Mixtral 8x22B: 15.1 (#293), Qwen2-72B: 21.2 (#275)
| Benchmark | Mixtral 8x22B | Qwen2-72B |
|---|---|---|
| GPQA Diamond | 34.1% | 40.8% |
| LMArena Expert | 1113 | 1171 |
| MMLU | 77.8% | 82.4% |
| MMLU-Pro | 46% | — |
| GPQA (HELM) | 33.4% | — |
Multilingual Qwen2-72B leads
Mixtral 8x22B: 32.8 (#255), Qwen2-72B: 35.9 (#244)
| Benchmark | Mixtral 8x22B | Qwen2-72B |
|---|---|---|
| LMArena Non-English | 1128 | 1176 |
| LMArena Chinese | 1116 | 1240 |
| LMArena French | 1166 | 1170 |
| LMArena German | 1141 | 1151 |
| LMArena Japanese | 1037 | 1111 |
| LMArena Korean | 1057 | 1083 |
| LMArena Russian | 1158 | 1169 |
| LMArena Spanish | 1151 | 1169 |
Instruction Following Qwen2-72B leads
Mixtral 8x22B: 57.7 (#266), Qwen2-72B: 61.7 (#241)
| Benchmark | Mixtral 8x22B | Qwen2-72B |
|---|---|---|
| LMArena Instruction Following | 1147 | 1181 |
| IFEval | 72.4% | — |
Long Context Qwen2-72B leads
Mixtral 8x22B: 34.7 (#247), Qwen2-72B: 36.1 (#235)
| Benchmark | Mixtral 8x22B | Qwen2-72B |
|---|---|---|
| LMArena Longer Query | 1144 | 1192 |
Writing & Preference Qwen2-72B leads
Mixtral 8x22B: 36.9 (#262), Qwen2-72B: 40.8 (#241)
| Benchmark | Mixtral 8x22B | Qwen2-72B |
|---|---|---|
| LMArena Text | 1162 | 1203 |
| LMArena Creative Writing | 1141 | 1181 |
| LMArena Multi-Turn | 1130 | 1196 |
| WildBench | 71.1% | — |
Frequently asked questions
Is Mixtral 8x22B better than Qwen2-72B?
Qwen2-72B is the stronger model overall, scoring 30.0 to 27.1 on the Noometry Index.
Is Mixtral 8x22B or Qwen2-72B better for coding?
Qwen2-72B scores higher on coding benchmarks: 29.1 versus 24.2 in the Noometry coding category.
How many benchmarks do Mixtral 8x22B and Qwen2-72B share?
24 benchmarks have published results for both models. Mixtral 8x22B has 34 scored results on Noometry and Qwen2-72B has 26.