Model comparison
GPT-4 Turbo vs Qwen3.8 Max
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 30.5 on the Noometry Index.
Last verified . 25 shared benchmarks.
Alibaba (Qwen)Summary
- They share 25 benchmarks with published results for both. GPT-4 Turbo scores higher in 0 categories and Qwen3.8 Max in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3.8 Max leads 73.2 to 9.0.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 6.7% for GPT-4 Turbo and 100% for Qwen3.8 Max.
- Qwen3.8 Max is cheaper at $2 / $6 per million input/output tokens, against $10 / $30 for GPT-4 Turbo.
- Qwen3.8 Max accepts more context: 1M tokens versus 128K.
Side by side
| GPT-4 Turbo | Qwen3.8 Max | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 30.5 | 56.8 |
| Released | 2023-11-06 | 2026-08-02 |
| Weights | Proprietary | Proprietary |
| Context window | 128K | 1M |
| Max output | 4K | 131K |
| Input $ / M tokens | $10 | $2 |
| Output $ / M tokens | $30 | $6 |
| Results tracked | 36 | 39 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.8 Max leads
GPT-4 Turbo: 33.8 (#249), Qwen3.8 Max: 53.5 (#29)
| Benchmark | GPT-4 Turbo | Qwen3.8 Max |
|---|---|---|
| LMArena Coding | 1268 | 1502 |
| DeepSWE | — | 57.5% |
| LMArena WebDev | — | 1674 |
| FrontierSWE | — | 17.8% |
| SciCode | — | 53.2% |
| WeirdML | 18% | — |
| BigCodeBench Instruct | 48.2% | — |
| BigCodeBench Complete | 58.2% | — |
| HumanEval+ | 86.6% | — |
| MBPP+ | 73.3% | — |
Agentic & Tool Use Not comparable
GPT-4 Turbo: —, Qwen3.8 Max: 45.4 (#14)
| Benchmark | GPT-4 Turbo | Qwen3.8 Max |
|---|---|---|
| APEX-Agents | — | 63.3% |
| τ²-bench Banking | — | 55.1% |
| GDP.pdf | — | 23.2% |
| METR Time Horizons | 36.7% | — |
Reasoning Qwen3.8 Max leads
GPT-4 Turbo: 15.3 (#317), Qwen3.8 Max: 54.4 (#26)
| Benchmark | GPT-4 Turbo | Qwen3.8 Max |
|---|---|---|
| Chess Puzzles | 6% | 40% |
| LMArena Hard Prompts | 1251 | 1496 |
| DTBench | 61.6% | 92% |
| LMCA | 9.8% | 46.2% |
| Epoch Capabilities Index | 127.25 | 156.41 |
| SimpleBench | 25.1% | — |
| NYT Connections (extended) | — | 88.3% |
| CritPt | — | 20% |
| Mystery Game Puzzles | — | 38% |
| ForecastBench | 59.4 | — |
Math Qwen3.8 Max leads
GPT-4 Turbo: 9.0 (#322), Qwen3.8 Max: 73.2 (#20)
| Benchmark | GPT-4 Turbo | Qwen3.8 Max |
|---|---|---|
| FrontierMath (Tiers 1-3) | 0.7% | 74.7% |
| OTIS Mock AIME 2024-2025 | 6.7% | 100% |
| LMArena Math | 1272 | 1499 |
| FrontierMath Tier 4 | — | 46.3% |
| ProofBench | — | 58% |
| MATH Level 5 | 46.7% | — |
Knowledge Qwen3.8 Max leads
GPT-4 Turbo: 24.3 (#268), Qwen3.8 Max: 61.7 (#27)
| Benchmark | GPT-4 Turbo | Qwen3.8 Max |
|---|---|---|
| GPQA Diamond | 46.6% | 92.7% |
| LMArena Expert | 1223 | 1507 |
| SimpleQA Verified | — | 47.3% |
| Confabulations | 28.4% | — |
| MMLU | 81.3% | — |
Multimodal Qwen3.8 Max leads
GPT-4 Turbo: 30.6 (#110), Qwen3.8 Max: 37.2 (#75)
| Benchmark | GPT-4 Turbo | Qwen3.8 Max |
|---|---|---|
| LMArena Vision | 1090 | 1314 |
| Furniture Assembly | — | 20% |
Multilingual Qwen3.8 Max leads
GPT-4 Turbo: 40.5 (#216), Qwen3.8 Max: 56.7 (#18)
| Benchmark | GPT-4 Turbo | Qwen3.8 Max |
|---|---|---|
| LMArena Non-English | 1245 | 1472 |
| LMArena Chinese | 1242 | 1538 |
| LMArena French | 1276 | 1503 |
| LMArena German | 1259 | 1483 |
| LMArena Japanese | 1194 | 1467 |
| LMArena Korean | 1187 | 1461 |
| LMArena Russian | 1259 | 1481 |
| LMArena Spanish | 1260 | 1492 |
Instruction Following Qwen3.8 Max leads
GPT-4 Turbo: 65.8 (#216), Qwen3.8 Max: 77.6 (#17)
| Benchmark | GPT-4 Turbo | Qwen3.8 Max |
|---|---|---|
| LMArena Instruction Following | 1249 | 1479 |
Long Context Qwen3.8 Max leads
GPT-4 Turbo: 38.0 (#206), Qwen3.8 Max: 45.6 (#31)
| Benchmark | GPT-4 Turbo | Qwen3.8 Max |
|---|---|---|
| LMArena Longer Query | 1254 | 1489 |
Writing & Preference Qwen3.8 Max leads
GPT-4 Turbo: 47.7 (#206), Qwen3.8 Max: 67.1 (#30)
| Benchmark | GPT-4 Turbo | Qwen3.8 Max |
|---|---|---|
| LMArena Text | 1272 | 1483 |
| LMArena Creative Writing | 1269 | 1479 |
| LMArena Multi-Turn | 1267 | 1489 |
Frequently asked questions
Is GPT-4 Turbo better than Qwen3.8 Max?
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 30.5 on the Noometry Index.
Which is cheaper, GPT-4 Turbo or Qwen3.8 Max?
Qwen3.8 Max is cheaper. It lists at $2 per million input tokens and $6 per million output tokens; GPT-4 Turbo lists at $10 and $30.
Is GPT-4 Turbo or Qwen3.8 Max better for coding?
Qwen3.8 Max scores higher on coding benchmarks: 53.5 versus 33.8 in the Noometry coding category.
Which has the bigger context window?
Qwen3.8 Max does, with 1M tokens against 128K.
How many benchmarks do GPT-4 Turbo and Qwen3.8 Max share?
25 benchmarks have published results for both models. GPT-4 Turbo has 36 scored results on Noometry and Qwen3.8 Max has 39.