Model comparison
Qwen2.5-Max vs Qwen3 235B-A22B
Qwen3 235B-A22B is the stronger model overall, scoring 43.5 to 40.7 on the Noometry Index.
Last verified . 20 shared benchmarks.
Alibaba (Qwen)Summary
- They share 20 benchmarks with published results for both. Qwen2.5-Max scores higher in 1 category and Qwen3 235B-A22B in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen3 235B-A22B leads 49.6 to 35.3.
- The biggest single-benchmark swing is Short-Story Creative Writing: 72.9% for Qwen2.5-Max and 83% for Qwen3 235B-A22B.
- Qwen3 235B-A22B has downloadable open weights; the other is API-only.
Side by side
| Qwen2.5-Max | Qwen3 235B-A22B | |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 40.7 | 43.5 |
| Released | 2025-01-25 | 2025-04 |
| Weights | Proprietary | Open |
| Context window | — | 131K |
| Max output | — | 16K |
| Input $ / M tokens | — | $0.70 |
| Output $ / M tokens | — | $2.80 |
| Results tracked | 27 | 49 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3 235B-A22B leads
Qwen2.5-Max: 41.8 (#117), Qwen3 235B-A22B: 44.3 (#75)
| Benchmark | Qwen2.5-Max | Qwen3 235B-A22B |
|---|---|---|
| LMArena Coding | 1359 | 1445 |
| Aider Polyglot | — | 59.6% |
| SciCode | — | 42.4% |
| WeirdML | — | 41% |
| LiveBench Coding | 64.4% | — |
Agentic & Tool Use Not comparable
Qwen2.5-Max: —, Qwen3 235B-A22B: 33.9 (#51)
| Benchmark | Qwen2.5-Max | Qwen3 235B-A22B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 52.1% |
| Vending-Bench 2 | — | -11.34 |
Reasoning Qwen2.5-Max leads
Qwen2.5-Max: 25.6 (#147), Qwen3 235B-A22B: 15.7 (#311)
| Benchmark | Qwen2.5-Max | Qwen3 235B-A22B |
|---|---|---|
| LMArena Hard Prompts | 1360 | 1433 |
| Epoch Capabilities Index | 132.53 | 143.85 |
| ARC-AGI-2 | — | 1.3% |
| SimpleBench | — | 31% |
| Kagi LLM Benchmark | — | 69.4% |
| ARC-AGI-1 | — | 11% |
| CritPt | — | 0% |
| Chess Puzzles | — | 12% |
| LiveBench Reasoning | 51.4% | — |
| Mystery Game Puzzles | — | 9% |
| DTBench | — | 80.3% |
| LiveBench Data Analysis | 67.9% | — |
| LMCA | — | 29.3% |
| ForecastBench | — | 59.7 |
| LiveBench | 62.3% | — |
Math Qwen3 235B-A22B leads
Qwen2.5-Max: 36.9 (#162), Qwen3 235B-A22B: 50.4 (#57)
| Benchmark | Qwen2.5-Max | Qwen3 235B-A22B |
|---|---|---|
| LMArena Math | 1369 | 1432 |
| OTIS Mock AIME 2024-2025 | — | 86.7% |
| Omni-MATH | — | 71.8% |
| LiveBench Math | 58.4% | — |
| MATH Level 5 | — | 68.9% |
| FrontierMath (Feb 2025 set) | — | 8.5% |
| FrontierMath Tier 4 (v1) | — | 0% |
Knowledge Qwen3 235B-A22B leads
Qwen2.5-Max: 35.3 (#186), Qwen3 235B-A22B: 49.6 (#73)
| Benchmark | Qwen2.5-Max | Qwen3 235B-A22B |
|---|---|---|
| Confabulations | 21.8% | 15.6% |
| LMArena Expert | 1337 | 1463 |
| GPQA Diamond | — | 80.1% |
| SimpleQA Verified | — | 40.4% |
| MMLU-Pro | — | 84.4% |
| Vectara Hallucination Rate | — | 9.3% |
| GPQA (HELM) | — | 72.7% |
Multilingual Qwen3 235B-A22B leads
Qwen2.5-Max: 48.1 (#146), Qwen3 235B-A22B: 52.3 (#89)
| Benchmark | Qwen2.5-Max | Qwen3 235B-A22B |
|---|---|---|
| LMArena Non-English | 1352 | 1409 |
| LMArena Chinese | 1382 | 1481 |
| LMArena French | 1396 | 1445 |
| LMArena German | 1350 | 1433 |
| LMArena Japanese | 1300 | 1399 |
| LMArena Korean | 1304 | 1391 |
| LMArena Russian | 1353 | 1411 |
| LMArena Spanish | 1377 | 1430 |
Instruction Following Qwen3 235B-A22B leads
Qwen2.5-Max: 71.3 (#152), Qwen3 235B-A22B: 72.6 (#136)
| Benchmark | Qwen2.5-Max | Qwen3 235B-A22B |
|---|---|---|
| LMArena Instruction Following | 1335 | 1408 |
| LiveBench Instruction Following | 75.3% | — |
| IFEval | — | 83.5% |
Long Context Qwen3 235B-A22B leads
Qwen2.5-Max: 41.4 (#142), Qwen3 235B-A22B: 46.1 (#26)
| Benchmark | Qwen2.5-Max | Qwen3 235B-A22B |
|---|---|---|
| LMArena Longer Query | 1358 | 1426 |
| Fiction.LiveBench | — | 75% |
Writing & Preference Qwen3 235B-A22B leads
Qwen2.5-Max: 55.4 (#146), Qwen3 235B-A22B: 59.6 (#108)
| Benchmark | Qwen2.5-Max | Qwen3 235B-A22B |
|---|---|---|
| LMArena Text | 1367 | 1419 |
| LMArena Creative Writing | 1339 | 1384 |
| Short-Story Creative Writing | 72.9% | 83% |
| LMArena Multi-Turn | 1364 | 1432 |
| EQ-Bench Creative Writing | — | 1366 |
| WildBench | — | 86.6% |
| LiveBench Language | 56.3% | — |
Frequently asked questions
Is Qwen2.5-Max better than Qwen3 235B-A22B?
Qwen3 235B-A22B is the stronger model overall, scoring 43.5 to 40.7 on the Noometry Index.
Is Qwen2.5-Max or Qwen3 235B-A22B better for coding?
Qwen3 235B-A22B scores higher on coding benchmarks: 44.3 versus 41.8 in the Noometry coding category.
How many benchmarks do Qwen2.5-Max and Qwen3 235B-A22B share?
20 benchmarks have published results for both models. Qwen2.5-Max has 27 scored results on Noometry and Qwen3 235B-A22B has 49.