# GPT-6 Astra vs Qwen3 Max

> GPT-6 Astra is the stronger model overall, scoring 70.8 to 43.7 on the Noometry Index. Qwen3 Max costs 8.3× less per token, which makes it the better buy when GPT-6 Astra's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/gpt-6-astra-vs-qwen3-max
- Last updated: 2026-10-12
- Shared benchmarks: 29

## Summary

- They share 29 benchmarks with published results for both. GPT-6 Astra scores higher in 8 categories and Qwen3 Max in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-6 Astra leads 85.1 to 22.6.
- The biggest single-benchmark swing is Mystery Game Puzzles: 84% for GPT-6 Astra and 5% for Qwen3 Max.
- Qwen3 Max is cheaper at $1.20 / $6 per million input/output tokens, against $10 / $50 for GPT-6 Astra.
- GPT-6 Astra accepts more context: 1.05M tokens versus 262K.

## Snapshot

| | GPT-6 Astra | Qwen3 Max |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 70.8 | 43.7 |
| Rank | 1 | 87 |
| Context | 1.05M | 262K |
| Input $/M | $10 | $1.20 |
| Output $/M | $50 | $6 |
| Weights | Proprietary | Proprietary |

## Coding

- GPT-6 Astra: 73.7 (#2)
- Qwen3 Max: 43.0 (#93)

| Benchmark | GPT-6 Astra | Qwen3 Max |
|---|---|---|
| LMArena Coding | 1487 | 1456 |
| ALE-Bench | 2,951 | 370.45 |
| DeepSWE | 74.1% | — |
| FrontierCode | 53.3% | — |
| LMArena WebDev | 1786 | — |
| FrontierSWE | 65.5% | — |
| SciCode | 56.5% | — |
| GSO | 79.4% | — |
| WeirdML | 93.6% | — |
| MirrorCode | 46.7% | — |

## Agentic & Tool Use

- GPT-6 Astra: 52.9 (#3)
- Qwen3 Max: —

| Benchmark | GPT-6 Astra | Qwen3 Max |
|---|---|---|
| Vending-Bench 2 | 15,515 | 71.56 |
| APEX-Agents | 64.7% | — |
| Remote Labor Index | 20.8% | — |
| BALROG | 68.3% | — |
| GDP.pdf | 34.2% | — |

## Reasoning

- GPT-6 Astra: 85.1 (#1)
- Qwen3 Max: 22.6 (#190)

| Benchmark | GPT-6 Astra | Qwen3 Max |
|---|---|---|
| NYT Connections (extended) | 98.1% | 30.1% |
| Chess Puzzles | 72% | 4% |
| LMArena Hard Prompts | 1462 | 1448 |
| Mystery Game Puzzles | 84% | 5% |
| DTBench | 97.3% | 82.1% |
| LMCA | 64.4% | 28.3% |
| Epoch Capabilities Index | 166.45 | 142.38 |
| ARC-AGI-2 | 95% | — |
| Kagi LLM Benchmark | — | 72.5% |
| ARC-AGI-1 | 98.5% | — |
| CritPt | 31.7% | — |
| EBR-Bench | 76.2% | — |
| Bench to the Future 3 | 0.14 | — |

## Math

- GPT-6 Astra: 93.5 (#2)
- Qwen3 Max: 38.7 (#131)

| Benchmark | GPT-6 Astra | Qwen3 Max |
|---|---|---|
| FrontierMath (Tiers 1-3) | 93.7% | 18.9% |
| OTIS Mock AIME 2024-2025 | 100% | 73.3% |
| LMArena Math | 1465 | 1446 |
| FrontierMath Tier 4 | 97.6% | — |
| ProofBench | 99% | — |
| MATH Level 5 | — | 97.1% |
| FrontierMath Erdős | 2.9% | — |

## Knowledge

- GPT-6 Astra: 75.3 (#1)
- Qwen3 Max: 48.1 (#78)

| Benchmark | GPT-6 Astra | Qwen3 Max |
|---|---|---|
| GPQA Diamond | 95.8% | 72.6% |
| SimpleQA Verified | 75.6% | 48.7% |
| LMArena Expert | 1483 | 1455 |
| Humanity's Last Exam | 54.8% | — |
| Vectara Hallucination Rate | 8.7% | — |

## Multimodal

- GPT-6 Astra: 55.0 (#3)
- Qwen3 Max: —

| Benchmark | GPT-6 Astra | Qwen3 Max |
|---|---|---|
| LMArena Vision | 1281 | — |
| Blueprint-Bench 2 | 49.7% | — |
| Furniture Assembly | 80% | — |
| LMArena Document | 1468 | — |

## Multilingual

- GPT-6 Astra: 53.7 (#61)
- Qwen3 Max: 53.7 (#62)

| Benchmark | GPT-6 Astra | Qwen3 Max |
|---|---|---|
| LMArena Non-English | 1430 | 1429 |
| LMArena Chinese | 1484 | 1478 |
| LMArena French | 1456 | 1449 |
| LMArena German | 1440 | 1463 |
| LMArena Japanese | 1379 | 1397 |
| LMArena Korean | 1426 | 1399 |
| LMArena Russian | 1436 | 1428 |
| LMArena Spanish | 1407 | 1462 |

## Instruction Following

- GPT-6 Astra: 76.3 (#44)
- Qwen3 Max: 74.8 (#87)

| Benchmark | GPT-6 Astra | Qwen3 Max |
|---|---|---|
| LMArena Instruction Following | 1450 | 1419 |

## Long Context

- GPT-6 Astra: 44.5 (#62)
- Qwen3 Max: 41.6 (#134)

| Benchmark | GPT-6 Astra | Qwen3 Max |
|---|---|---|
| LMArena Longer Query | 1456 | 1438 |
| Fiction.LiveBench | — | 66.7% |
| CL-bench | — | 14.5% |

## Writing & Preference

- GPT-6 Astra: 75.3 (#7)
- Qwen3 Max: 62.4 (#76)

| Benchmark | GPT-6 Astra | Qwen3 Max |
|---|---|---|
| LMArena Text | 1441 | 1439 |
| LMArena Creative Writing | 1418 | 1402 |
| LMArena Multi-Turn | 1448 | 1446 |
| EQ-Bench Creative Writing | 2173 | — |

## FAQ

### Is GPT-6 Astra better than Qwen3 Max?

GPT-6 Astra is the stronger model overall, scoring 70.8 to 43.7 on the Noometry Index. Qwen3 Max costs 8.3× less per token, which makes it the better buy when GPT-6 Astra's lead doesn't matter for your workload.

### Which is cheaper, GPT-6 Astra or Qwen3 Max?

Qwen3 Max is cheaper. It lists at $1.20 per million input tokens and $6 per million output tokens; GPT-6 Astra lists at $10 and $50.

### Is GPT-6 Astra or Qwen3 Max better for coding?

GPT-6 Astra scores higher on coding benchmarks: 73.7 versus 43.0 in the Noometry coding category.

### Which has the bigger context window?

GPT-6 Astra does, with 1.05M tokens against 262K.

### How many benchmarks do GPT-6 Astra and Qwen3 Max share?

29 benchmarks have published results for both models. GPT-6 Astra has 56 scored results on Noometry and Qwen3 Max has 33.