Model comparison
Mixtral 8x22B vs Qwen3.8 Max
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 27.1 on the Noometry Index.
Last verified . 20 shared benchmarks.
Mistral AI
Alibaba (Qwen)Summary
- They share 20 benchmarks with published results for both. Mixtral 8x22B scores higher in 0 categories and Qwen3.8 Max in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3.8 Max leads 73.2 to 22.9.
- The biggest single-benchmark swing is GPQA Diamond: 34.1% for Mixtral 8x22B and 92.7% for Qwen3.8 Max.
- Both cost about the same: $2 input and $6 output per million tokens.
- Qwen3.8 Max accepts more context: 1M tokens versus 64K.
- Mixtral 8x22B has downloadable open weights; the other is API-only.
Side by side
| Mixtral 8x22B | Qwen3.8 Max | |
|---|---|---|
| Provider | Mistral AI | Alibaba (Qwen) |
| Noometry Index | 27.1 | 56.8 |
| Released | 2024-04-17 | 2026-08-02 |
| Weights | Open | Proprietary |
| Context window | 64K | 1M |
| Max output | 64K | 131K |
| Input $ / M tokens | $2 | $2 |
| Output $ / M tokens | $6 | $6 |
| Results tracked | 34 | 39 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.8 Max leads
Mixtral 8x22B: 24.2 (#329), Qwen3.8 Max: 53.5 (#29)
| Benchmark | Mixtral 8x22B | Qwen3.8 Max |
|---|---|---|
| LMArena Coding | 1166 | 1502 |
| DeepSWE | — | 57.5% |
| LMArena WebDev | — | 1674 |
| FrontierSWE | — | 17.8% |
| SciCode | — | 53.2% |
| WeirdML | 3.2% | — |
| BigCodeBench Instruct | 40.6% | — |
| BigCodeBench Complete | 50.2% | — |
| HumanEval+ | 72% | — |
| MBPP+ | 64.3% | — |
Agentic & Tool Use Qwen3.8 Max leads
Mixtral 8x22B: 23.1 (#127), Qwen3.8 Max: 45.4 (#14)
| Benchmark | Mixtral 8x22B | Qwen3.8 Max |
|---|---|---|
| APEX-Agents | — | 63.3% |
| τ²-bench Banking | — | 55.1% |
| Cybench | 7.5% | — |
| GDP.pdf | — | 23.2% |
Reasoning Qwen3.8 Max leads
Mixtral 8x22B: 19.9 (#248), Qwen3.8 Max: 54.4 (#26)
| Benchmark | Mixtral 8x22B | Qwen3.8 Max |
|---|---|---|
| LMArena Hard Prompts | 1150 | 1496 |
| DTBench | 55.1% | 92% |
| Epoch Capabilities Index | 122.03 | 156.41 |
| NYT Connections (extended) | — | 88.3% |
| CritPt | — | 20% |
| Chess Puzzles | — | 40% |
| Mystery Game Puzzles | — | 38% |
| LMCA | — | 46.2% |
| ForecastBench | 56.3 | — |
Math Qwen3.8 Max leads
Mixtral 8x22B: 22.9 (#275), Qwen3.8 Max: 73.2 (#20)
| Benchmark | Mixtral 8x22B | Qwen3.8 Max |
|---|---|---|
| LMArena Math | 1184 | 1499 |
| FrontierMath (Tiers 1-3) | — | 74.7% |
| FrontierMath Tier 4 | — | 46.3% |
| OTIS Mock AIME 2024-2025 | — | 100% |
| ProofBench | — | 58% |
| Omni-MATH | 16.3% | — |
| MATH Level 5 | 24.2% | — |
Knowledge Qwen3.8 Max leads
Mixtral 8x22B: 15.1 (#293), Qwen3.8 Max: 61.7 (#27)
| Benchmark | Mixtral 8x22B | Qwen3.8 Max |
|---|---|---|
| GPQA Diamond | 34.1% | 92.7% |
| LMArena Expert | 1113 | 1507 |
| SimpleQA Verified | — | 47.3% |
| MMLU-Pro | 46% | — |
| GPQA (HELM) | 33.4% | — |
| MMLU | 77.8% | — |
Multimodal Not comparable
Mixtral 8x22B: —, Qwen3.8 Max: 37.2 (#75)
| Benchmark | Mixtral 8x22B | Qwen3.8 Max |
|---|---|---|
| LMArena Vision | — | 1314 |
| Furniture Assembly | — | 20% |
Multilingual Qwen3.8 Max leads
Mixtral 8x22B: 32.8 (#255), Qwen3.8 Max: 56.7 (#18)
| Benchmark | Mixtral 8x22B | Qwen3.8 Max |
|---|---|---|
| LMArena Non-English | 1128 | 1472 |
| LMArena Chinese | 1116 | 1538 |
| LMArena French | 1166 | 1503 |
| LMArena German | 1141 | 1483 |
| LMArena Japanese | 1037 | 1467 |
| LMArena Korean | 1057 | 1461 |
| LMArena Russian | 1158 | 1481 |
| LMArena Spanish | 1151 | 1492 |
Instruction Following Qwen3.8 Max leads
Mixtral 8x22B: 57.7 (#266), Qwen3.8 Max: 77.6 (#17)
| Benchmark | Mixtral 8x22B | Qwen3.8 Max |
|---|---|---|
| LMArena Instruction Following | 1147 | 1479 |
| IFEval | 72.4% | — |
Long Context Qwen3.8 Max leads
Mixtral 8x22B: 34.7 (#247), Qwen3.8 Max: 45.6 (#31)
| Benchmark | Mixtral 8x22B | Qwen3.8 Max |
|---|---|---|
| LMArena Longer Query | 1144 | 1489 |
Writing & Preference Qwen3.8 Max leads
Mixtral 8x22B: 36.9 (#262), Qwen3.8 Max: 67.1 (#30)
| Benchmark | Mixtral 8x22B | Qwen3.8 Max |
|---|---|---|
| LMArena Text | 1162 | 1483 |
| LMArena Creative Writing | 1141 | 1479 |
| LMArena Multi-Turn | 1130 | 1489 |
| WildBench | 71.1% | — |
Frequently asked questions
Is Mixtral 8x22B better than Qwen3.8 Max?
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 27.1 on the Noometry Index.
Which is cheaper, Mixtral 8x22B or Qwen3.8 Max?
Qwen3.8 Max is cheaper. It lists at $2 per million input tokens and $6 per million output tokens; Mixtral 8x22B lists at $2 and $6.
Is Mixtral 8x22B or Qwen3.8 Max better for coding?
Qwen3.8 Max scores higher on coding benchmarks: 53.5 versus 24.2 in the Noometry coding category.
Which has the bigger context window?
Qwen3.8 Max does, with 1M tokens against 64K.
How many benchmarks do Mixtral 8x22B and Qwen3.8 Max share?
20 benchmarks have published results for both models. Mixtral 8x22B has 34 scored results on Noometry and Qwen3.8 Max has 39.