Model comparison
GPT-5.6 Sol vs Qwen3 32B
GPT-5.6 Sol is the stronger model overall, scoring 64.4 to 38.8 on the Noometry Index. Qwen3 32B costs 6.5× less per token, which makes it the better buy when GPT-5.6 Sol's lead doesn't matter for your workload.
Last verified . 23 shared benchmarks.
Alibaba (Qwen)Summary
- They share 23 benchmarks with published results for both. GPT-5.6 Sol scores higher in 9 categories and Qwen3 32B in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-5.6 Sol leads 74.7 to 19.6.
- The biggest single-benchmark swing is Chess Puzzles: 64% for GPT-5.6 Sol and 5% for Qwen3 32B.
- Qwen3 32B is cheaper at $0.70 / $2.80 per million input/output tokens, against $4 / $20 for GPT-5.6 Sol.
- GPT-5.6 Sol accepts more context: 1.05M tokens versus 131K.
- Qwen3 32B has downloadable open weights; the other is API-only.
Side by side
| GPT-5.6 Sol | Qwen3 32B | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 64.4 | 38.8 |
| Released | 2026-07-09 | 2025-04 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 131K |
| Max output | 128K | 16K |
| Input $ / M tokens | $4 | $0.70 |
| Output $ / M tokens | $20 | $2.80 |
| Results tracked | 64 | 26 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.6 Sol leads
GPT-5.6 Sol: 61.1 (#7), Qwen3 32B: 37.3 (#184)
| Benchmark | GPT-5.6 Sol | Qwen3 32B |
|---|---|---|
| SciCode | 57.1% | 35.4% |
| LMArena Coding | 1493 | 1358 |
| DeepSWE | 72.7% | — |
| FrontierCode | 47.5% | — |
| Aider Polyglot | — | 40% |
| CursorBench | 41.7% | — |
| LMArena WebDev | 1619 | — |
| FrontierSWE | 32.2% | — |
| WeirdML | 89.4% | — |
| MirrorCode | 20% | — |
| ALE-Bench | 2,177 | — |
Agentic & Tool Use GPT-5.6 Sol leads
GPT-5.6 Sol: 50.2 (#7), Qwen3 32B: 32.3 (#58)
| Benchmark | GPT-5.6 Sol | Qwen3 32B |
|---|---|---|
| APEX-Agents | 51.4% | — |
| Berkeley Function Calling Leaderboard | — | 48.7% |
| OSWorld 2.0 | 27.3% | — |
| τ²-bench Banking | 46.9% | — |
| PostTrainBench | 36.2% | — |
| BALROG | 60% | — |
| GBAEval | 52.6% | — |
| GDP.pdf | 30.7% | — |
| LMArena Search | 1257 | — |
| Vending-Bench 2 | 9,619 | — |
Reasoning GPT-5.6 Sol leads
GPT-5.6 Sol: 74.7 (#6), Qwen3 32B: 19.6 (#233)
| Benchmark | GPT-5.6 Sol | Qwen3 32B |
|---|---|---|
| Kagi LLM Benchmark | 67% | 54.9% |
| CritPt | 32.3% | 0.3% |
| Chess Puzzles | 64% | 5% |
| LMArena Hard Prompts | 1481 | 1335 |
| DTBench | 96% | 67.5% |
| LMCA | 59.2% | 17.3% |
| Epoch Capabilities Index | 161.99 | 138.51 |
| ARC-AGI-2 | 92.5% | — |
| SimpleBench | 71.7% | — |
| NYT Connections (extended) | 93.8% | — |
| ARC-AGI-1 | 97.5% | — |
| EnigmaEval | 37.1% | — |
| EBR-Bench | 44.8% | — |
| Mystery Game Puzzles | 58% | — |
| Surface Evolver Bench | 93.1% | — |
| Bench to the Future 3 | 0.14 | — |
Math GPT-5.6 Sol leads
GPT-5.6 Sol: 85.5 (#7), Qwen3 32B: 37.9 (#89)
| Benchmark | GPT-5.6 Sol | Qwen3 32B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 100% | 66.9% |
| LMArena Math | 1472 | 1399 |
| FrontierMath (Tiers 1-3) | 89.1% | — |
| FrontierMath Tier 4 | 82.9% | — |
| ProofBench | 83% | — |
| FrontierMath Erdős | 0% | — |
Knowledge GPT-5.6 Sol leads
GPT-5.6 Sol: 64.3 (#16), Qwen3 32B: 40.0 (#112)
| Benchmark | GPT-5.6 Sol | Qwen3 32B |
|---|---|---|
| GPQA Diamond | 93.5% | 65.7% |
| Vectara Hallucination Rate | 12.4% | 5.9% |
| LMArena Expert | 1517 | 1362 |
| SimpleQA Verified | 69.7% | — |
Multimodal Not comparable
GPT-5.6 Sol: 48.9 (#8), Qwen3 32B: —
| Benchmark | GPT-5.6 Sol | Qwen3 32B |
|---|---|---|
| LMArena Vision | 1281 | — |
| Blueprint-Bench 2 | 33.6% | — |
| Furniture Assembly | 56.7% | — |
| LMArena Document | 1483 | — |
Multilingual GPT-5.6 Sol leads
GPT-5.6 Sol: 55.4 (#32), Qwen3 32B: 45.8 (#162)
| Benchmark | GPT-5.6 Sol | Qwen3 32B |
|---|---|---|
| LMArena Non-English | 1450 | 1317 |
| LMArena Chinese | 1523 | 1357 |
| LMArena German | 1476 | 1341 |
| LMArena Russian | 1466 | 1311 |
| LMArena French | 1475 | — |
| LMArena Japanese | 1475 | — |
| LMArena Korean | 1434 | — |
| LMArena Spanish | 1449 | — |
Instruction Following GPT-5.6 Sol leads
GPT-5.6 Sol: 77.8 (#16), Qwen3 32B: 69.1 (#174)
| Benchmark | GPT-5.6 Sol | Qwen3 32B |
|---|---|---|
| LMArena Instruction Following | 1480 | 1305 |
Long Context GPT-5.6 Sol leads
GPT-5.6 Sol: 45.3 (#42), Qwen3 32B: 43.7 (#83)
| Benchmark | GPT-5.6 Sol | Qwen3 32B |
|---|---|---|
| LMArena Longer Query | 1479 | 1327 |
| Fiction.LiveBench | — | 74.2% |
Writing & Preference GPT-5.6 Sol leads
GPT-5.6 Sol: 73.5 (#12), Qwen3 32B: 53.1 (#158)
| Benchmark | GPT-5.6 Sol | Qwen3 32B |
|---|---|---|
| LMArena Text | 1456 | 1340 |
| LMArena Creative Writing | 1450 | 1297 |
| LMArena Multi-Turn | 1456 | 1331 |
| EQ-Bench Creative Writing | 1972 | — |
| EQ-Bench 4 | 1250 | — |
Frequently asked questions
Is GPT-5.6 Sol better than Qwen3 32B?
GPT-5.6 Sol is the stronger model overall, scoring 64.4 to 38.8 on the Noometry Index. Qwen3 32B costs 6.5× less per token, which makes it the better buy when GPT-5.6 Sol's lead doesn't matter for your workload.
Which is cheaper, GPT-5.6 Sol or Qwen3 32B?
Qwen3 32B is cheaper. It lists at $0.70 per million input tokens and $2.80 per million output tokens; GPT-5.6 Sol lists at $4 and $20.
Is GPT-5.6 Sol or Qwen3 32B better for coding?
GPT-5.6 Sol scores higher on coding benchmarks: 61.1 versus 37.3 in the Noometry coding category.
Which has the bigger context window?
GPT-5.6 Sol does, with 1.05M tokens against 131K.
How many benchmarks do GPT-5.6 Sol and Qwen3 32B share?
23 benchmarks have published results for both models. GPT-5.6 Sol has 64 scored results on Noometry and Qwen3 32B has 26.