# GPT-4 vs Qwen3 Max

> Qwen3 Max is the stronger model overall, scoring 43.5 to 28.9 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gpt-4-vs-qwen3-max
- Last updated: 2026-10-10
- Shared benchmarks: 25

## Summary

- They share 25 benchmarks with published results for both. GPT-4 scores higher in 0 categories and Qwen3 Max in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen3 Max leads 48.1 to 18.4.
- The biggest single-benchmark swing is MATH Level 5: 23% for GPT-4 and 97.1% for Qwen3 Max.
- Qwen3 Max is cheaper at $1.20 / $6 per million input/output tokens, against $30 / $60 for GPT-4.
- Qwen3 Max accepts more context: 262K tokens versus 8K.

## Snapshot

| | GPT-4 | Qwen3 Max |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 28.9 | 43.5 |
| Rank | 308 | 82 |
| Context | 8K | 262K |
| Input $/M | $30 | $1.20 |
| Output $/M | $60 | $6 |
| Weights | Proprietary | Proprietary |

## Coding

- GPT-4: 31.2 (#277)
- Qwen3 Max: 42.4 (#93)

| Benchmark | GPT-4 | Qwen3 Max |
|---|---|---|
| LMArena Coding | 1254 | 1456 |
| WeirdML | 12.4% | — |
| BigCodeBench Instruct | 46% | — |
| BigCodeBench Complete | 57.2% | — |
| ALE-Bench | — | 370.45 |
| HumanEval+ | 79.3% | — |

## Agentic & Tool Use

- GPT-4: —
- Qwen3 Max: —

| Benchmark | GPT-4 | Qwen3 Max |
|---|---|---|
| METR Time Horizons | 36.1% | — |
| Vending-Bench 2 | — | 71.56 |

## Reasoning

- GPT-4: 17.3 (#285)
- Qwen3 Max: 22.0 (#186)

| Benchmark | GPT-4 | Qwen3 Max |
|---|---|---|
| Chess Puzzles | 4% | 4% |
| LMArena Hard Prompts | 1241 | 1448 |
| Mystery Game Puzzles | 12% | 5% |
| DTBench | 62.7% | 82.1% |
| LMCA | 17.1% | 28.3% |
| Epoch Capabilities Index | 125.88 | 142.41 |
| Kagi LLM Benchmark | — | 72.5% |
| NYT Connections (extended) | — | 30.1% |
| BIG-Bench Hard | 75.1% | — |
| ForecastBench | 57.8 | — |
| HellaSwag | 95.3% | — |
| WinoGrande | 87.5% | — |

## Math

- GPT-4: 10.1 (#304)
- Qwen3 Max: 37.6 (#94)

| Benchmark | GPT-4 | Qwen3 Max |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.1% | 73.3% |
| LMArena Math | 1269 | 1447 |
| MATH Level 5 | 23% | 97.1% |
| FrontierMath (Tiers 1-3) | — | 18.9% |
| GSM8K | 92% | — |

## Knowledge

- GPT-4: 18.4 (#277)
- Qwen3 Max: 48.1 (#72)

| Benchmark | GPT-4 | Qwen3 Max |
|---|---|---|
| GPQA Diamond | 35.7% | 72.6% |
| LMArena Expert | 1211 | 1457 |
| SimpleQA Verified | — | 48.7% |
| MMLU | 86.4% | — |
| TriviaQA | 84.8% | — |

## Multilingual

- GPT-4: 40.8 (#210)
- Qwen3 Max: 53.9 (#58)

| Benchmark | GPT-4 | Qwen3 Max |
|---|---|---|
| LMArena Non-English | 1246 | 1429 |
| LMArena Chinese | 1242 | 1479 |
| LMArena French | 1282 | 1450 |
| LMArena German | 1251 | 1463 |
| LMArena Japanese | 1209 | 1399 |
| LMArena Korean | 1184 | 1397 |
| LMArena Russian | 1251 | 1428 |
| LMArena Spanish | 1261 | 1463 |

## Instruction Following

- GPT-4: 65.5 (#217)
- Qwen3 Max: 75.0 (#81)

| Benchmark | GPT-4 | Qwen3 Max |
|---|---|---|
| LMArena Instruction Following | 1241 | 1420 |

## Long Context

- GPT-4: 37.7 (#207)
- Qwen3 Max: 41.6 (#129)

| Benchmark | GPT-4 | Qwen3 Max |
|---|---|---|
| LMArena Longer Query | 1244 | 1439 |
| Fiction.LiveBench | — | 66.7% |
| CL-bench | — | 14.5% |

## Writing & Preference

- GPT-4: 35.0 (#263)
- Qwen3 Max: 62.6 (#72)

| Benchmark | GPT-4 | Qwen3 Max |
|---|---|---|
| LMArena Text | 1263 | 1439 |
| LMArena Creative Writing | 1244 | 1401 |
| LMArena Multi-Turn | 1257 | 1446 |
| EQ-Bench Creative Writing | 752 | — |

## FAQ

### Is GPT-4 better than Qwen3 Max?

Qwen3 Max is the stronger model overall, scoring 43.5 to 28.9 on the Noometry Index.

### Which is cheaper, GPT-4 or Qwen3 Max?

Qwen3 Max is cheaper. It lists at $1.20 per million input tokens and $6 per million output tokens; GPT-4 lists at $30 and $60.

### Is GPT-4 or Qwen3 Max better for coding?

Qwen3 Max scores higher on coding benchmarks: 42.4 versus 31.2 in the Noometry coding category.

### Which has the bigger context window?

Qwen3 Max does, with 262K tokens against 8K.

### How many benchmarks do GPT-4 and Qwen3 Max share?

25 benchmarks have published results for both models. GPT-4 has 38 scored results on Noometry and Qwen3 Max has 33.