Model comparison
GPT-4.5 vs Qwen3 235B-A22B
Qwen3 235B-A22B is the stronger model overall, scoring 43.5 to 37.2 on the Noometry Index.
Last verified . 30 shared benchmarks.
Alibaba (Qwen)Summary
- They share 30 benchmarks with published results for both. GPT-4.5 scores higher in 2 categories and Qwen3 235B-A22B in 7 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3 235B-A22B leads 50.4 to 32.6.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 37.8% for GPT-4.5 and 86.7% for Qwen3 235B-A22B.
- Qwen3 235B-A22B has downloadable open weights; the other is API-only.
Side by side
| GPT-4.5 | Qwen3 235B-A22B | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 37.2 | 43.5 |
| Released | 2025-02-27 | 2025-04 |
| Weights | Proprietary | Open |
| Context window | — | 131K |
| Max output | — | 16K |
| Input $ / M tokens | — | $0.70 |
| Output $ / M tokens | — | $2.80 |
| Results tracked | 42 | 49 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3 235B-A22B leads
GPT-4.5: 42.2 (#109), Qwen3 235B-A22B: 44.3 (#75)
| Benchmark | GPT-4.5 | Qwen3 235B-A22B |
|---|---|---|
| Aider Polyglot | 44.9% | 59.6% |
| WeirdML | 39.4% | 41% |
| LMArena Coding | 1396 | 1445 |
| SciCode | — | 42.4% |
| LiveBench Coding | 75.2% | — |
Agentic & Tool Use Qwen3 235B-A22B leads
GPT-4.5: 27.9 (#97), Qwen3 235B-A22B: 33.9 (#51)
| Benchmark | GPT-4.5 | Qwen3 235B-A22B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 52.1% |
| Cybench | 17.5% | — |
| Vending-Bench 2 | — | -11.34 |
Reasoning Qwen3 235B-A22B leads
GPT-4.5: 13.9 (#330), Qwen3 235B-A22B: 15.7 (#311)
| Benchmark | GPT-4.5 | Qwen3 235B-A22B |
|---|---|---|
| ARC-AGI-2 | 0.8% | 1.3% |
| SimpleBench | 34.5% | 31% |
| ARC-AGI-1 | 10.3% | 11% |
| LMArena Hard Prompts | 1403 | 1433 |
| Epoch Capabilities Index | 136.74 | 143.85 |
| ForecastBench | 61.7 | 59.7 |
| Kagi LLM Benchmark | — | 69.4% |
| CritPt | — | 0% |
| Chess Puzzles | — | 12% |
| EnigmaEval | 3.2% | — |
| LiveBench Reasoning | 71.1% | — |
| Mystery Game Puzzles | — | 9% |
| DTBench | — | 80.3% |
| LiveBench Data Analysis | 64.3% | — |
| LMCA | — | 29.3% |
| LiveBench | 69% | — |
Math Qwen3 235B-A22B leads
GPT-4.5: 32.6 (#211), Qwen3 235B-A22B: 50.4 (#57)
| Benchmark | GPT-4.5 | Qwen3 235B-A22B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 37.8% | 86.7% |
| LMArena Math | 1412 | 1432 |
| MATH Level 5 | 78.6% | 68.9% |
| Omni-MATH | — | 71.8% |
| LiveBench Math | 69.3% | — |
| FrontierMath (Feb 2025 set) | — | 8.5% |
| FrontierMath Tier 4 (v1) | — | 0% |
Knowledge Qwen3 235B-A22B leads
GPT-4.5: 32.5 (#211), Qwen3 235B-A22B: 49.6 (#73)
| Benchmark | GPT-4.5 | Qwen3 235B-A22B |
|---|---|---|
| GPQA Diamond | 68.7% | 80.1% |
| Confabulations | 13.6% | 15.6% |
| LMArena Expert | 1394 | 1463 |
| Humanity's Last Exam | 5.4% | — |
| SimpleQA Verified | — | 40.4% |
| MMLU-Pro | — | 84.4% |
| Vectara Hallucination Rate | — | 9.3% |
| GPQA (HELM) | — | 72.7% |
Multimodal Not comparable
GPT-4.5: 37.6 (#71), Qwen3 235B-A22B: —
| Benchmark | GPT-4.5 | Qwen3 235B-A22B |
|---|---|---|
| LMArena Vision | 1195 | — |
| VPCT | 45% | — |
Multilingual Too close to call
GPT-4.5: 52.5 (#83), Qwen3 235B-A22B: 52.3 (#89)
| Benchmark | GPT-4.5 | Qwen3 235B-A22B |
|---|---|---|
| LMArena Non-English | 1413 | 1409 |
| LMArena Chinese | 1421 | 1481 |
| LMArena French | 1418 | 1445 |
| LMArena German | 1457 | 1433 |
| LMArena Japanese | 1416 | 1399 |
| LMArena Korean | 1392 | 1391 |
| LMArena Russian | 1419 | 1411 |
| LMArena Spanish | — | 1430 |
Instruction Following Too close to call
GPT-4.5: 72.6 (#134), Qwen3 235B-A22B: 72.6 (#136)
| Benchmark | GPT-4.5 | Qwen3 235B-A22B |
|---|---|---|
| LMArena Instruction Following | 1404 | 1408 |
| LiveBench Instruction Following | 72.3% | — |
| IFEval | — | 83.5% |
Long Context Qwen3 235B-A22B leads
GPT-4.5: 40.4 (#155), Qwen3 235B-A22B: 46.1 (#26)
| Benchmark | GPT-4.5 | Qwen3 235B-A22B |
|---|---|---|
| Fiction.LiveBench | 63.9% | 75% |
| LMArena Longer Query | 1406 | 1426 |
Writing & Preference Qwen3 235B-A22B leads
GPT-4.5: 56.9 (#134), Qwen3 235B-A22B: 59.6 (#108)
| Benchmark | GPT-4.5 | Qwen3 235B-A22B |
|---|---|---|
| LMArena Text | 1417 | 1419 |
| LMArena Creative Writing | 1394 | 1384 |
| Short-Story Creative Writing | 75.6% | 83% |
| EQ-Bench Creative Writing | 1258 | 1366 |
| LMArena Multi-Turn | 1444 | 1432 |
| WildBench | — | 86.6% |
| LiveBench Language | 61.5% | — |
Frequently asked questions
Is GPT-4.5 better than Qwen3 235B-A22B?
Qwen3 235B-A22B is the stronger model overall, scoring 43.5 to 37.2 on the Noometry Index.
Is GPT-4.5 or Qwen3 235B-A22B better for coding?
Qwen3 235B-A22B scores higher on coding benchmarks: 44.3 versus 42.2 in the Noometry coding category.
How many benchmarks do GPT-4.5 and Qwen3 235B-A22B share?
30 benchmarks have published results for both models. GPT-4.5 has 42 scored results on Noometry and Qwen3 235B-A22B has 49.