Model comparison
GPT-5 vs Qwen3 Max
GPT-5 is the stronger model overall, scoring 50.9 to 43.7 on the Noometry Index.
Last verified . 30 shared benchmarks.
Alibaba (Qwen)Summary
- They share 30 benchmarks with published results for both. GPT-5 scores higher in 6 categories and Qwen3 Max in 2 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in long context, where GPT-5 leads 69.5 to 41.6.
- The biggest single-benchmark swing is FrontierMath (Tiers 1-3): 55.4% for GPT-5 and 18.9% for Qwen3 Max.
- Qwen3 Max is cheaper at $1.20 / $6 per million input/output tokens, against $1.25 / $10 for GPT-5.
- GPT-5 accepts more context: 400K tokens versus 262K.
Side by side
| GPT-5 | Qwen3 Max | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 50.9 | 43.7 |
| Released | 2025-08-07 | 2025-09-23 |
| Weights | Proprietary | Proprietary |
| Context window | 400K | 262K |
| Max output | 128K | 66K |
| Input $ / M tokens | $1.25 | $1.20 |
| Output $ / M tokens | $10 | $6 |
| Results tracked | 69 | 33 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5 leads
GPT-5: 50.3 (#47), Qwen3 Max: 43.0 (#93)
| Benchmark | GPT-5 | Qwen3 Max |
|---|---|---|
| LMArena Coding | 1436 | 1456 |
| ALE-Bench | 1,162 | 370.45 |
| SWE-bench Verified | 73.6% | — |
| SWE-bench Verified (bash only) | 65% | — |
| Aider Polyglot | 88% | — |
| LMArena WebDev | 1418 | — |
| SciCode | 42.9% | — |
| GSO | 6.9% | — |
| WeirdML | 60.7% | — |
| AlgoTune | 1.67 | — |
Agentic & Tool Use Not comparable
GPT-5: 33.1 (#56), Qwen3 Max: —
| Benchmark | GPT-5 | Qwen3 Max |
|---|---|---|
| Terminal-Bench | 49.6% | — |
| GDPval | 34.8% | — |
| Remote Labor Index | 1.7% | — |
| DeepResearch Bench | 49.6% | — |
| BALROG | 32.8% | — |
| LMArena Search | 1133 | — |
| METR Time Horizons | 69.6% | — |
| Vending-Bench 2 | — | 71.56 |
Reasoning GPT-5 leads
GPT-5: 38.3 (#64), Qwen3 Max: 22.6 (#190)
| Benchmark | GPT-5 | Qwen3 Max |
|---|---|---|
| Kagi LLM Benchmark | 72.7% | 72.5% |
| Chess Puzzles | 37% | 4% |
| LMArena Hard Prompts | 1416 | 1448 |
| Mystery Game Puzzles | 23% | 5% |
| DTBench | 90.7% | 82.1% |
| LMCA | 40% | 28.3% |
| Epoch Capabilities Index | 150 | 142.38 |
| ARC-AGI-2 | 9.9% | — |
| SimpleBench | 56.7% | — |
| NYT Connections (extended) | — | 30.1% |
| ARC-AGI-1 | 65.7% | — |
| CritPt | 12.6% | — |
| EnigmaEval | 10.5% | — |
| EBR-Bench | 12.7% | — |
| ForecastBench | 61.4 | — |
Math GPT-5 leads
GPT-5: 55.0 (#44), Qwen3 Max: 38.7 (#131)
| Benchmark | GPT-5 | Qwen3 Max |
|---|---|---|
| FrontierMath (Tiers 1-3) | 55.4% | 18.9% |
| OTIS Mock AIME 2024-2025 | 91.4% | 73.3% |
| LMArena Math | 1407 | 1446 |
| MATH Level 5 | 98.1% | 97.1% |
| FrontierMath Tier 4 | 22% | — |
| ProofBench | 18% | — |
| Omni-MATH | 64.7% | — |
| FrontierMath (Feb 2025 set) | 32.4% | — |
| FrontierMath Tier 4 (v1) | 12.5% | — |
Knowledge GPT-5 leads
GPT-5: 56.6 (#43), Qwen3 Max: 48.1 (#78)
| Benchmark | GPT-5 | Qwen3 Max |
|---|---|---|
| GPQA Diamond | 86.2% | 72.6% |
| SimpleQA Verified | 50.1% | 48.7% |
| LMArena Expert | 1419 | 1455 |
| Humanity's Last Exam | 25.3% | — |
| MMLU-Pro | 86.3% | — |
| Confabulations | 10.3% | — |
| Vectara Hallucination Rate | 14.7% | — |
| GPQA (HELM) | 79.2% | — |
Multimodal Not comparable
GPT-5: 46.8 (#13), Qwen3 Max: —
| Benchmark | GPT-5 | Qwen3 Max |
|---|---|---|
| LMArena Vision | 1232 | — |
| GeoBench | 81% | — |
| VPCT | 66% | — |
Multilingual Qwen3 Max leads
GPT-5: 51.4 (#110), Qwen3 Max: 53.7 (#62)
| Benchmark | GPT-5 | Qwen3 Max |
|---|---|---|
| LMArena Non-English | 1397 | 1429 |
| LMArena Chinese | 1422 | 1478 |
| LMArena French | 1410 | 1449 |
| LMArena German | 1416 | 1463 |
| LMArena Japanese | 1409 | 1397 |
| LMArena Korean | 1360 | 1399 |
| LMArena Russian | 1406 | 1428 |
| LMArena Spanish | 1399 | 1462 |
Instruction Following Too close to call
GPT-5: 73.8 (#113), Qwen3 Max: 74.8 (#87)
| Benchmark | GPT-5 | Qwen3 Max |
|---|---|---|
| LMArena Instruction Following | 1388 | 1419 |
| IFEval | 87.5% | — |
Long Context GPT-5 leads
GPT-5: 69.5 (#2), Qwen3 Max: 41.6 (#134)
| Benchmark | GPT-5 | Qwen3 Max |
|---|---|---|
| Fiction.LiveBench | 97.2% | 66.7% |
| LMArena Longer Query | 1399 | 1438 |
| CL-bench | — | 14.5% |
Writing & Preference Too close to call
GPT-5: 63.4 (#65), Qwen3 Max: 62.4 (#76)
| Benchmark | GPT-5 | Qwen3 Max |
|---|---|---|
| LMArena Text | 1406 | 1439 |
| LMArena Creative Writing | 1365 | 1402 |
| LMArena Multi-Turn | 1426 | 1446 |
| Short-Story Creative Writing | 86% | — |
| EQ-Bench Creative Writing | 1627 | — |
| WildBench | 85.7% | — |
Frequently asked questions
Is GPT-5 better than Qwen3 Max?
GPT-5 is the stronger model overall, scoring 50.9 to 43.7 on the Noometry Index.
Which is cheaper, GPT-5 or Qwen3 Max?
Qwen3 Max is cheaper. It lists at $1.20 per million input tokens and $6 per million output tokens; GPT-5 lists at $1.25 and $10.
Is GPT-5 or Qwen3 Max better for coding?
GPT-5 scores higher on coding benchmarks: 50.3 versus 43.0 in the Noometry coding category.
Which has the bigger context window?
GPT-5 does, with 400K tokens against 262K.
How many benchmarks do GPT-5 and Qwen3 Max share?
30 benchmarks have published results for both models. GPT-5 has 69 scored results on Noometry and Qwen3 Max has 33.