Model comparison
GPT-5.6 Terra vs Qwen3.8 Max
GPT-5.6 Terra is the stronger model overall, scoring 59.2 to 56.8 on the Noometry Index. Qwen3.8 Max costs 1.5× less per token, which makes it the better buy when GPT-5.6 Terra's lead doesn't matter for your workload.
Last verified . 37 shared benchmarks.
Alibaba (Qwen)Summary
- They share 37 benchmarks with published results for both. GPT-5.6 Terra scores higher in 5 categories and Qwen3.8 Max in 5 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in multimodal, where GPT-5.6 Terra leads 47.3 to 37.2.
- The biggest single-benchmark swing is Furniture Assembly: 54.2% for GPT-5.6 Terra and 20% for Qwen3.8 Max.
- Qwen3.8 Max is cheaper at $2 / $6 per million input/output tokens, against $2 / $12 for GPT-5.6 Terra.
- GPT-5.6 Terra accepts more context: 1.05M tokens versus 1M.
Side by side
| GPT-5.6 Terra | Qwen3.8 Max | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 59.2 | 56.8 |
| Released | 2026-07-09 | 2026-08-02 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 1M |
| Max output | 128K | 131K |
| Input $ / M tokens | $2 | $2 |
| Output $ / M tokens | $12 | $6 |
| Results tracked | 52 | 39 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.6 Terra leads
GPT-5.6 Terra: 57.7 (#19), Qwen3.8 Max: 53.5 (#29)
| Benchmark | GPT-5.6 Terra | Qwen3.8 Max |
|---|---|---|
| DeepSWE | 69.6% | 57.5% |
| LMArena WebDev | 1522 | 1674 |
| SciCode | 55% | 53.2% |
| LMArena Coding | 1484 | 1502 |
| FrontierCode | 41.3% | — |
| CursorBench | 41.3% | — |
| FrontierSWE | — | 17.8% |
| WeirdML | 78.3% | — |
| ALE-Bench | 1,951 | — |
Agentic & Tool Use Qwen3.8 Max leads
GPT-5.6 Terra: 40.1 (#25), Qwen3.8 Max: 45.4 (#14)
| Benchmark | GPT-5.6 Terra | Qwen3.8 Max |
|---|---|---|
| APEX-Agents | 58.2% | 63.3% |
| GDP.pdf | 24.7% | 23.2% |
| τ²-bench Banking | — | 55.1% |
| BALROG | 53.2% | — |
| Vending-Bench 2 | 7,343 | — |
Reasoning GPT-5.6 Terra leads
GPT-5.6 Terra: 60.7 (#21), Qwen3.8 Max: 54.4 (#26)
| Benchmark | GPT-5.6 Terra | Qwen3.8 Max |
|---|---|---|
| NYT Connections (extended) | 78.4% | 88.3% |
| CritPt | 30% | 20% |
| Chess Puzzles | 54% | 40% |
| LMArena Hard Prompts | 1468 | 1496 |
| Mystery Game Puzzles | 35% | 38% |
| DTBench | 93.3% | 92% |
| LMCA | 55% | 46.2% |
| Epoch Capabilities Index | 159.62 | 156.41 |
| ARC-AGI-2 | 83.9% | — |
| SimpleBench | 48.9% | — |
| Kagi LLM Benchmark | 51.3% | — |
| ARC-AGI-1 | 96.5% | — |
| Surface Evolver Bench | 83.8% | — |
Math GPT-5.6 Terra leads
GPT-5.6 Terra: 81.6 (#12), Qwen3.8 Max: 73.2 (#20)
| Benchmark | GPT-5.6 Terra | Qwen3.8 Max |
|---|---|---|
| FrontierMath (Tiers 1-3) | 86% | 74.7% |
| FrontierMath Tier 4 | 70.7% | 46.3% |
| OTIS Mock AIME 2024-2025 | 99.7% | 100% |
| ProofBench | 74% | 58% |
| LMArena Math | 1466 | 1499 |
Knowledge Too close to call
GPT-5.6 Terra: 61.2 (#30), Qwen3.8 Max: 61.7 (#27)
| Benchmark | GPT-5.6 Terra | Qwen3.8 Max |
|---|---|---|
| GPQA Diamond | 93.3% | 92.7% |
| SimpleQA Verified | 43.2% | 47.3% |
| LMArena Expert | 1492 | 1507 |
Multimodal GPT-5.6 Terra leads
GPT-5.6 Terra: 47.3 (#11), Qwen3.8 Max: 37.2 (#75)
| Benchmark | GPT-5.6 Terra | Qwen3.8 Max |
|---|---|---|
| LMArena Vision | 1271 | 1314 |
| Furniture Assembly | 54.2% | 20% |
| Blueprint-Bench 2 | 30.8% | — |
| LMArena Document | 1472 | — |
Multilingual Qwen3.8 Max leads
GPT-5.6 Terra: 54.4 (#44), Qwen3.8 Max: 56.7 (#18)
| Benchmark | GPT-5.6 Terra | Qwen3.8 Max |
|---|---|---|
| LMArena Non-English | 1439 | 1472 |
| LMArena Chinese | 1513 | 1538 |
| LMArena French | 1471 | 1503 |
| LMArena German | 1460 | 1483 |
| LMArena Japanese | 1457 | 1467 |
| LMArena Korean | 1425 | 1461 |
| LMArena Russian | 1450 | 1481 |
| LMArena Spanish | 1448 | 1492 |
Instruction Following Qwen3.8 Max leads
GPT-5.6 Terra: 76.4 (#40), Qwen3.8 Max: 77.6 (#17)
| Benchmark | GPT-5.6 Terra | Qwen3.8 Max |
|---|---|---|
| LMArena Instruction Following | 1454 | 1479 |
Long Context Qwen3.8 Max leads
GPT-5.6 Terra: 44.4 (#68), Qwen3.8 Max: 45.6 (#31)
| Benchmark | GPT-5.6 Terra | Qwen3.8 Max |
|---|---|---|
| LMArena Longer Query | 1451 | 1489 |
Writing & Preference GPT-5.6 Terra leads
GPT-5.6 Terra: 70.2 (#23), Qwen3.8 Max: 67.1 (#30)
| Benchmark | GPT-5.6 Terra | Qwen3.8 Max |
|---|---|---|
| LMArena Text | 1447 | 1483 |
| LMArena Creative Writing | 1410 | 1479 |
| LMArena Multi-Turn | 1449 | 1489 |
| EQ-Bench Creative Writing | 1855 | — |
| EQ-Bench 4 | 1234 | — |
Frequently asked questions
Is GPT-5.6 Terra better than Qwen3.8 Max?
GPT-5.6 Terra is the stronger model overall, scoring 59.2 to 56.8 on the Noometry Index. Qwen3.8 Max costs 1.5× less per token, which makes it the better buy when GPT-5.6 Terra's lead doesn't matter for your workload.
Which is cheaper, GPT-5.6 Terra or Qwen3.8 Max?
Qwen3.8 Max is cheaper. It lists at $2 per million input tokens and $6 per million output tokens; GPT-5.6 Terra lists at $2 and $12.
Is GPT-5.6 Terra or Qwen3.8 Max better for coding?
GPT-5.6 Terra scores higher on coding benchmarks: 57.7 versus 53.5 in the Noometry coding category.
Which has the bigger context window?
GPT-5.6 Terra does, with 1.05M tokens against 1M.
How many benchmarks do GPT-5.6 Terra and Qwen3.8 Max share?
37 benchmarks have published results for both models. GPT-5.6 Terra has 52 scored results on Noometry and Qwen3.8 Max has 39.