Model comparison
GPT-5 vs Qwen3-4B
GPT-5 is the stronger model overall, scoring 50.9 to 31.9 on the Noometry Index.
Last verified . 4 shared benchmarks.
Alibaba (Qwen)Summary
- They share 4 benchmarks with published results for both. GPT-5 scores higher in 4 categories and Qwen3-4B in 0 categories; 4 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5 leads 55.0 to 29.7.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 91.4% for GPT-5 and 52.2% for Qwen3-4B.
- Qwen3-4B has downloadable open weights; the other is API-only.
Side by side
| GPT-5 | Qwen3-4B | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 50.9 | 31.9 |
| Released | 2025-08-07 | 2025-04-29 |
| Weights | Proprietary | Open |
| Context window | 400K | — |
| Max output | 128K | — |
| Input $ / M tokens | $1.25 | — |
| Output $ / M tokens | $10 | — |
| Results tracked | 69 | 6 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Not comparable
GPT-5: 50.3 (#47), Qwen3-4B: —
| Benchmark | GPT-5 | Qwen3-4B |
|---|---|---|
| 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% | — |
| LMArena Coding | 1436 | — |
| ALE-Bench | 1,162 | — |
| AlgoTune | 1.67 | — |
Agentic & Tool Use GPT-5 leads
GPT-5: 33.1 (#56), Qwen3-4B: 27.6 (#100)
| Benchmark | GPT-5 | Qwen3-4B |
|---|---|---|
| Terminal-Bench | 49.6% | — |
| Berkeley Function Calling Leaderboard | — | 35.7% |
| GDPval | 34.8% | — |
| Remote Labor Index | 1.7% | — |
| DeepResearch Bench | 49.6% | — |
| BALROG | 32.8% | — |
| LMArena Search | 1133 | — |
| METR Time Horizons | 69.6% | — |
Reasoning GPT-5 leads
GPT-5: 38.3 (#64), Qwen3-4B: 19.2 (#268)
| Benchmark | GPT-5 | Qwen3-4B |
|---|---|---|
| Chess Puzzles | 37% | 4% |
| ARC-AGI-2 | 9.9% | — |
| SimpleBench | 56.7% | — |
| Kagi LLM Benchmark | 72.7% | — |
| ARC-AGI-1 | 65.7% | — |
| CritPt | 12.6% | — |
| EnigmaEval | 10.5% | — |
| EBR-Bench | 12.7% | — |
| LMArena Hard Prompts | 1416 | — |
| Mystery Game Puzzles | 23% | — |
| DTBench | 90.7% | — |
| LMCA | 40% | — |
| Epoch Capabilities Index | 150 | — |
| ForecastBench | 61.4 | — |
Math GPT-5 leads
GPT-5: 55.0 (#44), Qwen3-4B: 29.7 (#240)
| Benchmark | GPT-5 | Qwen3-4B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 91.4% | 52.2% |
| FrontierMath (Tiers 1-3) | 55.4% | — |
| FrontierMath Tier 4 | 22% | — |
| MathArena Final-Answer Competitions | — | 38.5% |
| ProofBench | 18% | — |
| Omni-MATH | 64.7% | — |
| LMArena Math | 1407 | — |
| MATH Level 5 | 98.1% | — |
| FrontierMath (Feb 2025 set) | 32.4% | — |
| FrontierMath Tier 4 (v1) | 12.5% | — |
Knowledge GPT-5 leads
GPT-5: 56.6 (#43), Qwen3-4B: 33.0 (#208)
| Benchmark | GPT-5 | Qwen3-4B |
|---|---|---|
| GPQA Diamond | 86.2% | 52.3% |
| Vectara Hallucination Rate | 14.7% | 5.7% |
| Humanity's Last Exam | 25.3% | — |
| SimpleQA Verified | 50.1% | — |
| MMLU-Pro | 86.3% | — |
| Confabulations | 10.3% | — |
| GPQA (HELM) | 79.2% | — |
| LMArena Expert | 1419 | — |
Multimodal Not comparable
GPT-5: 46.8 (#13), Qwen3-4B: —
| Benchmark | GPT-5 | Qwen3-4B |
|---|---|---|
| LMArena Vision | 1232 | — |
| GeoBench | 81% | — |
| VPCT | 66% | — |
Multilingual Not comparable
GPT-5: 51.4 (#110), Qwen3-4B: —
| Benchmark | GPT-5 | Qwen3-4B |
|---|---|---|
| LMArena Non-English | 1397 | — |
| LMArena Chinese | 1422 | — |
| LMArena French | 1410 | — |
| LMArena German | 1416 | — |
| LMArena Japanese | 1409 | — |
| LMArena Korean | 1360 | — |
| LMArena Russian | 1406 | — |
| LMArena Spanish | 1399 | — |
Instruction Following Not comparable
GPT-5: 73.8 (#113), Qwen3-4B: —
| Benchmark | GPT-5 | Qwen3-4B |
|---|---|---|
| IFEval | 87.5% | — |
| LMArena Instruction Following | 1388 | — |
Long Context Not comparable
GPT-5: 69.5 (#2), Qwen3-4B: —
| Benchmark | GPT-5 | Qwen3-4B |
|---|---|---|
| Fiction.LiveBench | 97.2% | — |
| LMArena Longer Query | 1399 | — |
Writing & Preference Not comparable
GPT-5: 63.4 (#65), Qwen3-4B: —
| Benchmark | GPT-5 | Qwen3-4B |
|---|---|---|
| LMArena Text | 1406 | — |
| LMArena Creative Writing | 1365 | — |
| Short-Story Creative Writing | 86% | — |
| EQ-Bench Creative Writing | 1627 | — |
| WildBench | 85.7% | — |
| LMArena Multi-Turn | 1426 | — |
Frequently asked questions
Is GPT-5 better than Qwen3-4B?
GPT-5 is the stronger model overall, scoring 50.9 to 31.9 on the Noometry Index.
How many benchmarks do GPT-5 and Qwen3-4B share?
4 benchmarks have published results for both models. GPT-5 has 69 scored results on Noometry and Qwen3-4B has 6.