Model comparison
GPT-5 Mini vs Qwen3 235B-A22B
Qwen3 235B-A22B is the stronger model overall, scoring 43.2 to 41.7 on the Noometry Index. GPT-5 Mini costs 1.8× less per token, which makes it the better buy when Qwen3 235B-A22B's lead doesn't matter for your workload.
Last verified . 47 shared benchmarks.
Alibaba (Qwen)Summary
- They share 47 benchmarks with published results for both. GPT-5 Mini scores higher in 2 categories and Qwen3 235B-A22B in 7 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-5 Mini leads 24.1 to 15.8.
- The biggest single-benchmark swing is ARC-AGI-1: 54.3% for GPT-5 Mini and 11% for Qwen3 235B-A22B.
- GPT-5 Mini is cheaper at $0.25 / $2 per million input/output tokens, against $0.70 / $2.80 for Qwen3 235B-A22B.
- GPT-5 Mini accepts more context: 400K tokens versus 131K.
- Qwen3 235B-A22B has downloadable open weights; the other is API-only.
Side by side
| GPT-5 Mini | Qwen3 235B-A22B | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 41.7 | 43.2 |
| Released | 2025-08-07 | 2025-04 |
| Weights | Proprietary | Open |
| Context window | 400K | 131K |
| Max output | 128K | 16K |
| Input $ / M tokens | $0.25 | $0.70 |
| Output $ / M tokens | $2 | $2.80 |
| Results tracked | 60 | 49 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3 235B-A22B leads
GPT-5 Mini: 40.0 (#134), Qwen3 235B-A22B: 43.8 (#72)
| Benchmark | GPT-5 Mini | Qwen3 235B-A22B |
|---|---|---|
| SciCode | 39.2% | 42.4% |
| WeirdML | 52.7% | 41% |
| LMArena Coding | 1406 | 1445 |
| SWE-bench Verified | 64.7% | — |
| SWE-bench Verified (bash only) | 59.8% | — |
| Aider Polyglot | — | 59.6% |
| SWE-bench Multilingual | 39.7% | — |
| ALE-Bench | 799.77 | — |
| AlgoTune | 1.38 | — |
Agentic & Tool Use Qwen3 235B-A22B leads
GPT-5 Mini: 31.0 (#64), Qwen3 235B-A22B: 33.7 (#48)
| Benchmark | GPT-5 Mini | Qwen3 235B-A22B |
|---|---|---|
| Berkeley Function Calling Leaderboard | 55.5% | 52.1% |
| Vending-Bench 2 | -31.18 | -11.34 |
| Terminal-Bench | 34.8% | — |
Reasoning GPT-5 Mini leads
GPT-5 Mini: 24.1 (#153), Qwen3 235B-A22B: 15.8 (#302)
| Benchmark | GPT-5 Mini | Qwen3 235B-A22B |
|---|---|---|
| ARC-AGI-2 | 4.4% | 1.3% |
| Kagi LLM Benchmark | 70.3% | 69.4% |
| ARC-AGI-1 | 54.3% | 11% |
| CritPt | 0% | 0% |
| Chess Puzzles | 30% | 12% |
| LMArena Hard Prompts | 1380 | 1433 |
| Mystery Game Puzzles | 10% | 9% |
| DTBench | 80.5% | 80.3% |
| LMCA | 34.2% | 29.3% |
| Epoch Capabilities Index | 145.51 | 143.87 |
| ForecastBench | 61 | 59.7 |
| SimpleBench | — | 31% |
| EnigmaEval | 8.2% | — |
Math Qwen3 235B-A22B leads
GPT-5 Mini: 46.3 (#59), Qwen3 235B-A22B: 48.5 (#51)
| Benchmark | GPT-5 Mini | Qwen3 235B-A22B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 86.7% | 86.7% |
| Omni-MATH | 72.2% | 71.8% |
| LMArena Math | 1377 | 1431 |
| MATH Level 5 | 97.8% | 68.9% |
| FrontierMath (Feb 2025 set) | 27.2% | 8.5% |
| FrontierMath Tier 4 (v1) | 6.3% | 0% |
| FrontierMath (Tiers 1-3) | 46.7% | — |
| FrontierMath Tier 4 | 12.2% | — |
| ProofBench | 9% | — |
Knowledge Qwen3 235B-A22B leads
GPT-5 Mini: 45.5 (#80), Qwen3 235B-A22B: 49.5 (#67)
| Benchmark | GPT-5 Mini | Qwen3 235B-A22B |
|---|---|---|
| GPQA Diamond | 75% | 80.1% |
| SimpleQA Verified | 21.6% | 40.4% |
| MMLU-Pro | 83.5% | 84.4% |
| Confabulations | 13.3% | 15.6% |
| Vectara Hallucination Rate | 12.9% | 9.3% |
| GPQA (HELM) | 75.6% | 72.7% |
| LMArena Expert | 1377 | 1460 |
| Humanity's Last Exam | 19.4% | — |
Multimodal Not comparable
GPT-5 Mini: 35.0 (#80), Qwen3 235B-A22B: —
| Benchmark | GPT-5 Mini | Qwen3 235B-A22B |
|---|---|---|
| LMArena Vision | 1202 | — |
| VPCT | 40.2% | — |
Multilingual Qwen3 235B-A22B leads
GPT-5 Mini: 49.1 (#131), Qwen3 235B-A22B: 52.5 (#84)
| Benchmark | GPT-5 Mini | Qwen3 235B-A22B |
|---|---|---|
| LMArena Non-English | 1363 | 1410 |
| LMArena Chinese | 1387 | 1481 |
| LMArena French | 1385 | 1445 |
| LMArena German | 1366 | 1432 |
| LMArena Japanese | 1342 | 1400 |
| LMArena Korean | 1306 | 1391 |
| LMArena Russian | 1362 | 1411 |
| LMArena Spanish | 1355 | 1429 |
Instruction Following GPT-5 Mini leads
GPT-5 Mini: 76.4 (#43), Qwen3 235B-A22B: 72.8 (#132)
| Benchmark | GPT-5 Mini | Qwen3 235B-A22B |
|---|---|---|
| IFEval | 92.7% | 83.5% |
| LMArena Instruction Following | 1357 | 1408 |
Long Context Qwen3 235B-A22B leads
GPT-5 Mini: 41.9 (#127), Qwen3 235B-A22B: 46.1 (#25)
| Benchmark | GPT-5 Mini | Qwen3 235B-A22B |
|---|---|---|
| Fiction.LiveBench | 69.4% | 75% |
| LMArena Longer Query | 1355 | 1426 |
Writing & Preference Qwen3 235B-A22B leads
GPT-5 Mini: 55.4 (#143), Qwen3 235B-A22B: 59.8 (#103)
| Benchmark | GPT-5 Mini | Qwen3 235B-A22B |
|---|---|---|
| LMArena Text | 1373 | 1419 |
| LMArena Creative Writing | 1325 | 1384 |
| Short-Story Creative Writing | 83.1% | 83% |
| EQ-Bench Creative Writing | 1313 | 1366 |
| WildBench | 85.5% | 86.6% |
| LMArena Multi-Turn | 1363 | 1432 |
Frequently asked questions
Is GPT-5 Mini better than Qwen3 235B-A22B?
Qwen3 235B-A22B is the stronger model overall, scoring 43.2 to 41.7 on the Noometry Index. GPT-5 Mini costs 1.8× less per token, which makes it the better buy when Qwen3 235B-A22B's lead doesn't matter for your workload.
Which is cheaper, GPT-5 Mini or Qwen3 235B-A22B?
GPT-5 Mini is cheaper. It lists at $0.25 per million input tokens and $2 per million output tokens; Qwen3 235B-A22B lists at $0.70 and $2.80.
Is GPT-5 Mini or Qwen3 235B-A22B better for coding?
Qwen3 235B-A22B scores higher on coding benchmarks: 43.8 versus 40.0 in the Noometry coding category.
Which has the bigger context window?
GPT-5 Mini does, with 400K tokens against 131K.
How many benchmarks do GPT-5 Mini and Qwen3 235B-A22B share?
47 benchmarks have published results for both models. GPT-5 Mini has 60 scored results on Noometry and Qwen3 235B-A22B has 49.