Model comparison
GPT-4o vs Qwen Max
Qwen Max is the stronger model overall, scoring 34.4 to 28.5 on the Noometry Index.
Last verified . 23 shared benchmarks.
Alibaba (Qwen)Summary
- They share 23 benchmarks with published results for both. GPT-4o scores higher in 3 categories and Qwen Max in 5 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Qwen Max leads 24.1 to 9.4.
- The biggest single-benchmark swing is Aider Polyglot: 45.3% for GPT-4o and 21.8% for Qwen Max.
- Qwen Max is cheaper at $1.60 / $6.40 per million input/output tokens, against $2.50 / $10 for GPT-4o.
- GPT-4o accepts more context: 128K tokens versus 33K.
Side by side
| GPT-4o | Qwen Max | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 28.5 | 34.4 |
| Released | 2024-05-13 | 2024-04-03 |
| Weights | Proprietary | Proprietary |
| Context window | 128K | 33K |
| Max output | 16K | 8K |
| Input $ / M tokens | $2.50 | $1.60 |
| Output $ / M tokens | $10 | $6.40 |
| Results tracked | 72 | 23 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen Max leads
GPT-4o: 25.1 (#323), Qwen Max: 30.3 (#288)
| Benchmark | GPT-4o | Qwen Max |
|---|---|---|
| Aider Polyglot | 45.3% | 21.8% |
| LMArena Coding | 1297 | 1288 |
| SWE-bench Verified | 31% | — |
| SWE-bench Verified (bash only) | 21.6% | — |
| GSO | 0% | — |
| WeirdML | 25.1% | — |
| BigCodeBench Instruct | 51.1% | — |
| LiveBench Coding | 51.4% | — |
| BigCodeBench Complete | 61.1% | — |
| CadEval | 26% | — |
| HumanEval+ | 87.2% | — |
| MBPP+ | 72.2% | — |
Agentic & Tool Use Not comparable
GPT-4o: 20.8 (#133), Qwen Max: —
| Benchmark | GPT-4o | Qwen Max |
|---|---|---|
| GDPval | 9.9% | — |
| TheAgentCompany | 8.6% | — |
| Cybench | 12.5% | — |
| BALROG | 32.3% | — |
| LMArena Search | 1006 | — |
| METR Time Horizons | 40.8% | — |
Reasoning Qwen Max leads
GPT-4o: 9.4 (#338), Qwen Max: 24.1 (#154)
| Benchmark | GPT-4o | Qwen Max |
|---|---|---|
| LMArena Hard Prompts | 1281 | 1270 |
| ARC-AGI-2 | 0% | — |
| SimpleBench | 17.8% | — |
| ARC-AGI-1 | 4.5% | — |
| CritPt | 0% | — |
| Chess Puzzles | 13% | — |
| EnigmaEval | 0.8% | — |
| LiveBench Reasoning | 55.8% | — |
| DTBench | 64.5% | — |
| LiveBench Data Analysis | 60.9% | — |
| LMCA | 16.6% | — |
| Epoch Capabilities Index | 128.97 | — |
| ForecastBench | 57.7 | — |
| LiveBench | 55.3% | — |
Math Qwen Max leads
GPT-4o: 10.2 (#303), Qwen Max: 21.0 (#272)
| Benchmark | GPT-4o | Qwen Max |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 6.4% | 16.1% |
| LMArena Math | 1285 | 1275 |
| MATH Level 5 | 53.3% | 67.2% |
| FrontierMath (Feb 2025 set) | 0.3% | 1% |
| FrontierMath (Tiers 1-3) | 0.4% | — |
| Omni-MATH | 29.3% | — |
| LiveBench Math | 49.5% | — |
Knowledge Qwen Max leads
GPT-4o: 28.8 (#235), Qwen Max: 30.3 (#222)
| Benchmark | GPT-4o | Qwen Max |
|---|---|---|
| GPQA Diamond | 49.2% | 56.1% |
| LMArena Expert | 1250 | 1248 |
| Humanity's Last Exam | 2.7% | — |
| SimpleQA Verified | 26% | — |
| MMLU-Pro | 71.3% | — |
| Confabulations | 15.3% | — |
| Vectara Hallucination Rate | 9.6% | — |
| GPQA (HELM) | 52% | — |
| MMLU | 88.1% | — |
Multimodal Not comparable
GPT-4o: 33.9 (#85), Qwen Max: —
| Benchmark | GPT-4o | Qwen Max |
|---|---|---|
| LMArena Vision | 1137 | — |
| Video-MME | 71.9% | — |
| GeoBench | 71% | — |
| VPCT | 40% | — |
| ScienceQA | 88.5% | — |
Multilingual GPT-4o leads
GPT-4o: 43.4 (#181), Qwen Max: 42.0 (#197)
| Benchmark | GPT-4o | Qwen Max |
|---|---|---|
| LMArena Non-English | 1283 | 1263 |
| LMArena Chinese | 1277 | 1254 |
| LMArena French | 1303 | 1329 |
| LMArena German | 1282 | 1254 |
| LMArena Japanese | 1257 | 1205 |
| LMArena Korean | 1234 | 1141 |
| LMArena Russian | 1286 | 1274 |
| LMArena Spanish | 1292 | 1290 |
Instruction Following Too close to call
GPT-4o: 66.8 (#202), Qwen Max: 66.7 (#203)
| Benchmark | GPT-4o | Qwen Max |
|---|---|---|
| LMArena Instruction Following | 1278 | 1262 |
| LiveBench Instruction Following | 68.6% | — |
| IFEval | 81.7% | — |
Long Context Too close to call
GPT-4o: 39.4 (#174), Qwen Max: 39.4 (#175)
| Benchmark | GPT-4o | Qwen Max |
|---|---|---|
| Fiction.LiveBench | 66.7% | 66.7% |
| LMArena Longer Query | 1289 | 1289 |
Writing & Preference GPT-4o leads
GPT-4o: 52.8 (#160), Qwen Max: 48.0 (#200)
| Benchmark | GPT-4o | Qwen Max |
|---|---|---|
| LMArena Text | 1300 | 1282 |
| LMArena Creative Writing | 1291 | 1248 |
| LMArena Multi-Turn | 1302 | 1277 |
| Short-Story Creative Writing | 81.8% | — |
| WildBench | 82.8% | — |
| LiveBench Language | 47.6% | — |
Frequently asked questions
Is GPT-4o better than Qwen Max?
Qwen Max is the stronger model overall, scoring 34.4 to 28.5 on the Noometry Index.
Which is cheaper, GPT-4o or Qwen Max?
Qwen Max is cheaper. It lists at $1.60 per million input tokens and $6.40 per million output tokens; GPT-4o lists at $2.50 and $10.
Is GPT-4o or Qwen Max better for coding?
Qwen Max scores higher on coding benchmarks: 30.3 versus 25.1 in the Noometry coding category.
Which has the bigger context window?
GPT-4o does, with 128K tokens against 33K.
How many benchmarks do GPT-4o and Qwen Max share?
23 benchmarks have published results for both models. GPT-4o has 72 scored results on Noometry and Qwen Max has 23.