Model comparison
Mistral Medium vs Qwen3 235B-A22B
Qwen3 235B-A22B is the stronger model overall, scoring 43.5 to 36.3 on the Noometry Index.
Last verified . 30 shared benchmarks.
Mistral AI
Alibaba (Qwen)Summary
- They share 30 benchmarks with published results for both. Mistral Medium scores higher in 3 categories and Qwen3 235B-A22B in 6 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen3 235B-A22B leads 49.6 to 25.0.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 32.2% for Mistral Medium and 86.7% for Qwen3 235B-A22B.
- Qwen3 235B-A22B is cheaper at $0.70 / $2.80 per million input/output tokens, against $1.50 / $7.50 for Mistral Medium.
- Mistral Medium accepts more context: 262K tokens versus 131K.
Side by side
| Mistral Medium | Qwen3 235B-A22B | |
|---|---|---|
| Provider | Mistral AI | Alibaba (Qwen) |
| Noometry Index | 36.3 | 43.5 |
| Released | 2023-12-11 | 2025-04 |
| Weights | Open | Open |
| Context window | 262K | 131K |
| Max output | 262K | 16K |
| Input $ / M tokens | $1.50 | $0.70 |
| Output $ / M tokens | $7.50 | $2.80 |
| Results tracked | 36 | 49 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3 235B-A22B leads
Mistral Medium: 34.2 (#243), Qwen3 235B-A22B: 44.3 (#75)
| Benchmark | Mistral Medium | Qwen3 235B-A22B |
|---|---|---|
| SciCode | 40.2% | 42.4% |
| WeirdML | 43.7% | 41% |
| LMArena Coding | 1434 | 1445 |
| FrontierCode | 8% | — |
| Aider Polyglot | — | 59.6% |
| ALE-Bench | 763.98 | — |
Agentic & Tool Use Qwen3 235B-A22B leads
Mistral Medium: 28.3 (#90), Qwen3 235B-A22B: 33.9 (#51)
| Benchmark | Mistral Medium | Qwen3 235B-A22B |
|---|---|---|
| Berkeley Function Calling Leaderboard | 37.7% | 52.1% |
| Vending-Bench 2 | — | -11.34 |
Reasoning Mistral Medium leads
Mistral Medium: 24.0 (#167), Qwen3 235B-A22B: 15.7 (#311)
| Benchmark | Mistral Medium | Qwen3 235B-A22B |
|---|---|---|
| Kagi LLM Benchmark | 50% | 69.4% |
| CritPt | 0% | 0% |
| LMArena Hard Prompts | 1426 | 1433 |
| DTBench | 75.5% | 80.3% |
| LMCA | 26.1% | 29.3% |
| ARC-AGI-2 | — | 1.3% |
| SimpleBench | — | 31% |
| ARC-AGI-1 | — | 11% |
| Chess Puzzles | — | 12% |
| Mystery Game Puzzles | — | 9% |
| Surface Evolver Bench | 26.9% | — |
| Epoch Capabilities Index | — | 143.85 |
| ForecastBench | — | 59.7 |
Math Qwen3 235B-A22B leads
Mistral Medium: 28.1 (#245), Qwen3 235B-A22B: 50.4 (#57)
| Benchmark | Mistral Medium | Qwen3 235B-A22B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 32.2% | 86.7% |
| LMArena Math | 1408 | 1432 |
| MATH Level 5 | 81.6% | 68.9% |
| FrontierMath (Feb 2025 set) | 0.3% | 8.5% |
| ProofBench | 9% | — |
| Omni-MATH | — | 71.8% |
| FrontierMath Tier 4 (v1) | — | 0% |
Knowledge Qwen3 235B-A22B leads
Mistral Medium: 25.0 (#265), Qwen3 235B-A22B: 49.6 (#73)
| Benchmark | Mistral Medium | Qwen3 235B-A22B |
|---|---|---|
| GPQA Diamond | 59.5% | 80.1% |
| Vectara Hallucination Rate | 22.7% | 9.3% |
| LMArena Expert | 1408 | 1463 |
| Humanity's Last Exam | 4.5% | — |
| SimpleQA Verified | — | 40.4% |
| MMLU-Pro | — | 84.4% |
| Confabulations | — | 15.6% |
| GPQA (HELM) | — | 72.7% |
Multimodal Not comparable
Mistral Medium: 35.3 (#88), Qwen3 235B-A22B: —
| Benchmark | Mistral Medium | Qwen3 235B-A22B |
|---|---|---|
| LMArena Vision | 1172 | — |
Multilingual Too close to call
Mistral Medium: 52.1 (#91), Qwen3 235B-A22B: 52.3 (#89)
| Benchmark | Mistral Medium | Qwen3 235B-A22B |
|---|---|---|
| LMArena Non-English | 1408 | 1409 |
| LMArena Chinese | 1447 | 1481 |
| LMArena French | 1459 | 1445 |
| LMArena German | 1432 | 1433 |
| LMArena Japanese | 1378 | 1399 |
| LMArena Korean | 1380 | 1391 |
| LMArena Russian | 1411 | 1411 |
| LMArena Spanish | 1433 | 1430 |
Instruction Following Mistral Medium leads
Mistral Medium: 73.7 (#116), Qwen3 235B-A22B: 72.6 (#136)
| Benchmark | Mistral Medium | Qwen3 235B-A22B |
|---|---|---|
| LMArena Instruction Following | 1398 | 1408 |
| IFEval | — | 83.5% |
Long Context Qwen3 235B-A22B leads
Mistral Medium: 42.9 (#114), Qwen3 235B-A22B: 46.1 (#26)
| Benchmark | Mistral Medium | Qwen3 235B-A22B |
|---|---|---|
| LMArena Longer Query | 1406 | 1426 |
| Fiction.LiveBench | — | 75% |
Writing & Preference Too close to call
Mistral Medium: 60.0 (#103), Qwen3 235B-A22B: 59.6 (#108)
| Benchmark | Mistral Medium | Qwen3 235B-A22B |
|---|---|---|
| LMArena Text | 1424 | 1419 |
| LMArena Creative Writing | 1391 | 1384 |
| Short-Story Creative Writing | 77.3% | 83% |
| LMArena Multi-Turn | 1418 | 1432 |
| EQ-Bench Creative Writing | — | 1366 |
| WildBench | — | 86.6% |
Frequently asked questions
Is Mistral Medium better than Qwen3 235B-A22B?
Qwen3 235B-A22B is the stronger model overall, scoring 43.5 to 36.3 on the Noometry Index.
Which is cheaper, Mistral Medium or Qwen3 235B-A22B?
Qwen3 235B-A22B is cheaper. It lists at $0.70 per million input tokens and $2.80 per million output tokens; Mistral Medium lists at $1.50 and $7.50.
Is Mistral Medium or Qwen3 235B-A22B better for coding?
Qwen3 235B-A22B scores higher on coding benchmarks: 44.3 versus 34.2 in the Noometry coding category.
Which has the bigger context window?
Mistral Medium does, with 262K tokens against 131K.
How many benchmarks do Mistral Medium and Qwen3 235B-A22B share?
30 benchmarks have published results for both models. Mistral Medium has 36 scored results on Noometry and Qwen3 235B-A22B has 49.