Model comparison
GPT-5.6 Sol vs Qwen3.7 Max
GPT-5.6 Sol is the stronger model overall, scoring 65.0 to 51.5 on the Noometry Index. Qwen3.7 Max costs 2.1× less per token, which makes it the better buy when GPT-5.6 Sol's lead doesn't matter for your workload.
Last verified . 32 shared benchmarks.
Alibaba (Qwen)Summary
- They share 32 benchmarks with published results for both. GPT-5.6 Sol scores higher in 7 categories and Qwen3.7 Max in 2 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in agentic & tool use, where GPT-5.6 Sol leads 50.3 to 22.1.
- The biggest single-benchmark swing is ProofBench: 83% for GPT-5.6 Sol and 26% for Qwen3.7 Max.
- Qwen3.7 Max is cheaper at $2.50 / $7.50 per million input/output tokens, against $4 / $20 for GPT-5.6 Sol.
- GPT-5.6 Sol accepts more context: 1.05M tokens versus 1M.
Side by side
| GPT-5.6 Sol | Qwen3.7 Max | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 65.0 | 51.5 |
| Released | 2026-07-09 | 2026-05-19 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 1M |
| Max output | 128K | 131K |
| Input $ / M tokens | $4 | $2.50 |
| Output $ / M tokens | $20 | $7.50 |
| Results tracked | 65 | 33 |
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: 65.1 (#7), Qwen3.7 Max: 50.4 (#45)
| Benchmark | GPT-5.6 Sol | Qwen3.7 Max |
|---|---|---|
| LMArena WebDev | 1618 | 1515 |
| SciCode | 57.1% | 48.8% |
| LMArena Coding | 1498 | 1498 |
| ALE-Bench | 2,177 | 1,189 |
| SWE-bench Verified | — | 77.3% |
| DeepSWE | 72.7% | — |
| FrontierCode | 47.5% | — |
| CursorBench | 41.7% | — |
| FrontierSWE | 32.2% | — |
| GSO | 76.5% | — |
| WeirdML | 89.4% | — |
| MirrorCode | 20% | — |
Agentic & Tool Use GPT-5.6 Sol leads
GPT-5.6 Sol: 50.3 (#7), Qwen3.7 Max: 22.1 (#135)
| Benchmark | GPT-5.6 Sol | Qwen3.7 Max |
|---|---|---|
| GBAEval | 52.6% | 0.4% |
| APEX-Agents | 51.4% | — |
| OSWorld 2.0 | 27.3% | — |
| τ²-bench Banking | 46.9% | — |
| PostTrainBench | 36.2% | — |
| BALROG | 60% | — |
| GDP.pdf | 30.7% | — |
| LMArena Search | 1257 | — |
| Vending-Bench 2 | 9,619 | — |
Reasoning GPT-5.6 Sol leads
GPT-5.6 Sol: 74.8 (#8), Qwen3.7 Max: 49.2 (#38)
| Benchmark | GPT-5.6 Sol | Qwen3.7 Max |
|---|---|---|
| SimpleBench | 71.7% | 70.4% |
| NYT Connections (extended) | 93.8% | 85.1% |
| CritPt | 32.3% | 13.4% |
| Chess Puzzles | 64% | 19% |
| EBR-Bench | 44.8% | 9.5% |
| LMArena Hard Prompts | 1484 | 1483 |
| Mystery Game Puzzles | 58% | 32% |
| DTBench | 96% | 92.3% |
| LMCA | 59.2% | 44% |
| Epoch Capabilities Index | 161.66 | 153.68 |
| ARC-AGI-2 | 92.5% | — |
| Kagi LLM Benchmark | 67% | — |
| ARC-AGI-1 | 97.5% | — |
| EnigmaEval | 37.1% | — |
| Surface Evolver Bench | 93.1% | — |
| Bench to the Future 3 | 0.14 | — |
Math GPT-5.6 Sol leads
GPT-5.6 Sol: 85.6 (#9), Qwen3.7 Max: 62.4 (#32)
| Benchmark | GPT-5.6 Sol | Qwen3.7 Max |
|---|---|---|
| FrontierMath (Tiers 1-3) | 89.1% | 64.6% |
| FrontierMath Tier 4 | 82.9% | 34.1% |
| OTIS Mock AIME 2024-2025 | 100% | 95.6% |
| ProofBench | 83% | 26% |
| LMArena Math | 1474 | 1490 |
| FrontierMath Erdős | 0% | — |
Knowledge GPT-5.6 Sol leads
GPT-5.6 Sol: 64.3 (#18), Qwen3.7 Max: 61.6 (#28)
| Benchmark | GPT-5.6 Sol | Qwen3.7 Max |
|---|---|---|
| GPQA Diamond | 93.5% | 90.9% |
| SimpleQA Verified | 69.7% | 55.8% |
| LMArena Expert | 1516 | 1488 |
| Vectara Hallucination Rate | 12.4% | — |
Multimodal Not comparable
GPT-5.6 Sol: 48.6 (#9), Qwen3.7 Max: —
| Benchmark | GPT-5.6 Sol | Qwen3.7 Max |
|---|---|---|
| LMArena Vision | 1281 | — |
| Blueprint-Bench 2 | 33.6% | — |
| Furniture Assembly | 56.7% | — |
| LMArena Document | 1483 | — |
Multilingual Qwen3.7 Max leads
GPT-5.6 Sol: 55.3 (#32), Qwen3.7 Max: 56.9 (#15)
| Benchmark | GPT-5.6 Sol | Qwen3.7 Max |
|---|---|---|
| LMArena Non-English | 1452 | 1474 |
| LMArena Chinese | 1527 | 1530 |
| LMArena Russian | 1468 | 1484 |
| LMArena French | 1477 | — |
| LMArena German | 1476 | — |
| LMArena Japanese | 1471 | — |
| LMArena Korean | 1442 | — |
| LMArena Spanish | 1441 | — |
Instruction Following Too close to call
GPT-5.6 Sol: 77.7 (#16), Qwen3.7 Max: 76.7 (#38)
| Benchmark | GPT-5.6 Sol | Qwen3.7 Max |
|---|---|---|
| LMArena Instruction Following | 1482 | 1460 |
Long Context Too close to call
GPT-5.6 Sol: 45.4 (#42), Qwen3.7 Max: 45.4 (#40)
| Benchmark | GPT-5.6 Sol | Qwen3.7 Max |
|---|---|---|
| LMArena Longer Query | 1480 | 1482 |
Writing & Preference GPT-5.6 Sol leads
GPT-5.6 Sol: 73.3 (#12), Qwen3.7 Max: 65.0 (#54)
| Benchmark | GPT-5.6 Sol | Qwen3.7 Max |
|---|---|---|
| LMArena Text | 1457 | 1476 |
| LMArena Creative Writing | 1448 | 1449 |
| EQ-Bench 4 | 1250 | 1110 |
| LMArena Multi-Turn | 1460 | 1481 |
| EQ-Bench Creative Writing | 1972 | — |
Frequently asked questions
Is GPT-5.6 Sol better than Qwen3.7 Max?
GPT-5.6 Sol is the stronger model overall, scoring 65.0 to 51.5 on the Noometry Index. Qwen3.7 Max costs 2.1× 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 Qwen3.7 Max?
Qwen3.7 Max is cheaper. It lists at $2.50 per million input tokens and $7.50 per million output tokens; GPT-5.6 Sol lists at $4 and $20.
Is GPT-5.6 Sol or Qwen3.7 Max better for coding?
GPT-5.6 Sol scores higher on coding benchmarks: 65.1 versus 50.4 in the Noometry coding category.
Which has the bigger context window?
GPT-5.6 Sol does, with 1.05M tokens against 1M.
How many benchmarks do GPT-5.6 Sol and Qwen3.7 Max share?
32 benchmarks have published results for both models. GPT-5.6 Sol has 65 scored results on Noometry and Qwen3.7 Max has 33.