Model comparison
GPT-5.4 Pro vs Qwen3.8 Max
GPT-5.4 Pro is the stronger model overall, scoring 58.9 to 56.8 on the Noometry Index. Qwen3.8 Max costs 23× less per token, which makes it the better buy when GPT-5.4 Pro's lead doesn't matter for your workload.
Last verified . 7 shared benchmarks.
Alibaba (Qwen)Summary
- They share 7 benchmarks with published results for both. GPT-5.4 Pro scores higher in 2 categories and Qwen3.8 Max in 2 categories; 3 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-5.4 Pro leads 70.7 to 54.4.
- The biggest single-benchmark swing is Chess Puzzles: 58.6% for GPT-5.4 Pro and 40% for Qwen3.8 Max.
- Qwen3.8 Max is cheaper at $2 / $6 per million input/output tokens, against $30 / $180 for GPT-5.4 Pro.
- GPT-5.4 Pro accepts more context: 1.05M tokens versus 1M.
Side by side
| GPT-5.4 Pro | Qwen3.8 Max | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 58.9 | 56.8 |
| Released | 2026-03-05 | 2026-08-02 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 1M |
| Max output | 128K | 131K |
| Input $ / M tokens | $30 | $2 |
| Output $ / M tokens | $180 | $6 |
| Results tracked | 16 | 39 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.8 Max leads
GPT-5.4 Pro: 43.2 (#90), Qwen3.8 Max: 53.5 (#29)
| Benchmark | GPT-5.4 Pro | Qwen3.8 Max |
|---|---|---|
| DeepSWE | — | 57.5% |
| LMArena WebDev | — | 1674 |
| FrontierSWE | — | 17.8% |
| SciCode | — | 53.2% |
| WeirdML | 57.4% | — |
| LMArena Coding | — | 1502 |
Agentic & Tool Use Not comparable
GPT-5.4 Pro: —, Qwen3.8 Max: 45.4 (#14)
| Benchmark | GPT-5.4 Pro | Qwen3.8 Max |
|---|---|---|
| APEX-Agents | — | 63.3% |
| τ²-bench Banking | — | 55.1% |
| GDP.pdf | — | 23.2% |
Reasoning GPT-5.4 Pro leads
GPT-5.4 Pro: 70.7 (#13), Qwen3.8 Max: 54.4 (#26)
| Benchmark | GPT-5.4 Pro | Qwen3.8 Max |
|---|---|---|
| CritPt | 30% | 20% |
| Chess Puzzles | 58.6% | 40% |
| Epoch Capabilities Index | 158.93 | 156.41 |
| ARC-AGI-2 | 83.3% | — |
| SimpleBench | 74.1% | — |
| NYT Connections (extended) | — | 88.3% |
| ARC-AGI-1 | 94.5% | — |
| EnigmaEval | 23.8% | — |
| LMArena Hard Prompts | — | 1496 |
| Mystery Game Puzzles | — | 38% |
| DTBench | — | 92% |
| LMCA | — | 46.2% |
Math Too close to call
GPT-5.4 Pro: 72.4 (#22), Qwen3.8 Max: 73.2 (#20)
| Benchmark | GPT-5.4 Pro | Qwen3.8 Max |
|---|---|---|
| FrontierMath (Tiers 1-3) | 82.5% | 74.7% |
| FrontierMath Tier 4 | 58.5% | 46.3% |
| OTIS Mock AIME 2024-2025 | — | 100% |
| ProofBench | — | 58% |
| LMArena Math | — | 1499 |
| FrontierMath (Feb 2025 set) | 50% | — |
| FrontierMath Tier 4 (v1) | 37.5% | — |
Knowledge GPT-5.4 Pro leads
GPT-5.4 Pro: 68.3 (#7), Qwen3.8 Max: 61.7 (#27)
| Benchmark | GPT-5.4 Pro | Qwen3.8 Max |
|---|---|---|
| GPQA Diamond | 94.6% | 92.7% |
| SimpleQA Verified | 46.3% | 47.3% |
| Humanity's Last Exam | 44.3% | — |
| Vectara Hallucination Rate | 8.3% | — |
| LMArena Expert | — | 1507 |
Multimodal Not comparable
GPT-5.4 Pro: —, Qwen3.8 Max: 37.2 (#75)
| Benchmark | GPT-5.4 Pro | Qwen3.8 Max |
|---|---|---|
| LMArena Vision | — | 1314 |
| Furniture Assembly | — | 20% |
Multilingual Not comparable
GPT-5.4 Pro: —, Qwen3.8 Max: 56.7 (#18)
| Benchmark | GPT-5.4 Pro | Qwen3.8 Max |
|---|---|---|
| LMArena Non-English | — | 1472 |
| LMArena Chinese | — | 1538 |
| LMArena French | — | 1503 |
| LMArena German | — | 1483 |
| LMArena Japanese | — | 1467 |
| LMArena Korean | — | 1461 |
| LMArena Russian | — | 1481 |
| LMArena Spanish | — | 1492 |
Instruction Following Not comparable
GPT-5.4 Pro: —, Qwen3.8 Max: 77.6 (#17)
| Benchmark | GPT-5.4 Pro | Qwen3.8 Max |
|---|---|---|
| LMArena Instruction Following | — | 1479 |
Long Context Not comparable
GPT-5.4 Pro: —, Qwen3.8 Max: 45.6 (#31)
| Benchmark | GPT-5.4 Pro | Qwen3.8 Max |
|---|---|---|
| LMArena Longer Query | — | 1489 |
Writing & Preference Not comparable
GPT-5.4 Pro: —, Qwen3.8 Max: 67.1 (#30)
| Benchmark | GPT-5.4 Pro | Qwen3.8 Max |
|---|---|---|
| LMArena Text | — | 1483 |
| LMArena Creative Writing | — | 1479 |
| LMArena Multi-Turn | — | 1489 |
Frequently asked questions
Is GPT-5.4 Pro better than Qwen3.8 Max?
GPT-5.4 Pro is the stronger model overall, scoring 58.9 to 56.8 on the Noometry Index. Qwen3.8 Max costs 23× less per token, which makes it the better buy when GPT-5.4 Pro's lead doesn't matter for your workload.
Which is cheaper, GPT-5.4 Pro or Qwen3.8 Max?
Qwen3.8 Max is cheaper. It lists at $2 per million input tokens and $6 per million output tokens; GPT-5.4 Pro lists at $30 and $180.
Is GPT-5.4 Pro or Qwen3.8 Max better for coding?
Qwen3.8 Max scores higher on coding benchmarks: 53.5 versus 43.2 in the Noometry coding category.
Which has the bigger context window?
GPT-5.4 Pro does, with 1.05M tokens against 1M.
How many benchmarks do GPT-5.4 Pro and Qwen3.8 Max share?
7 benchmarks have published results for both models. GPT-5.4 Pro has 16 scored results on Noometry and Qwen3.8 Max has 39.