GPT-6 Astra vs Qwen3 8
GPT-6 Astra 
Qwen3 8 Verdict
- GPT-6 Astra (max) leads on quality: 70.6 vs 63.9.
- GPT-6 Astra (max) is stronger in agents & tools, coding, composite, knowledge, maths, reasoning.
- Qwen3 8 (max) is stronger in human preference, long context, multimodal.
- Qwen3 8 (max) is 8.3× cheaper ($2.67 vs $22.00 per 1M blended).
- GPT-6 Astra (max) streams 1.4× faster (54 vs 39 tokens per second).
| Metric | GPT-6 Astra (max) | Qwen3 8 (max) |
|---|---|---|
| BenchLeader Index | 70.6 | 63.9 |
| Agents & tools score | 68.2 | 56.1 |
| Coding score | 79.0 | 71.7 |
| Composite score | 73.1 | 70.8 |
| Knowledge score | 79.8 | 62.3 |
| Maths score | 77.0 | – |
| Reasoning score | 70.9 | 67.7 |
| Human preference score | – | 67.9 |
| Long context score | – | 65.7 |
| Multimodal score | – | 67.2 |
| Blended price $/M | $22.00 | $2.67 |
| Output speed | 54 tok/s | 39 tok/s |
| Time to first answer | 328.6 s | 53.2 s |
| Context window | 1.1M | 1M |
| GPQA Diamond | 95.8% | – |
| FrontierMath Tiers 1–3 | 93.7% | – |
| FrontierMath Tier 4 | 97.6% | – |
| OTIS Mock AIME | 100.0% | – |
| SimpleQA Verified | 75.6% | – |
| Terminal-Bench | 58.2% | – |
| SciCode | 56.5% | – |
| FrontierCode | 53.3% | – |
| Epoch Capabilities Index | – | 156.6 |
| LMArena Text | – | 1480 |
| LMArena Hard Prompts | – | 1502 |
| LMArena Coding | – | 1520 |
| LMArena WebDev | 1796 | 1685 |
| LMArena Vision | – | 1313 |
| LMArena Agent | 12.5 | 3.9 |
| LiveBench | 82.2% | 78.5% |
| LiveBench Reasoning | 92.7% | 88.2% |
| LiveBench Coding | 80.4% | 72.9% |
| LiveBench Agentic Coding | 57.3% | 64.7% |
| LiveBench Mathematics | 96.8% | 91.3% |
| LiveBench Data Analysis | 83.0% | 78.4% |
| LiveBench Language | 89.4% | 79.7% |
| AA Intelligence Index | – | 40.3 |
| AA-LCR | – | 78.3% |
| MMMU-Pro | – | 82.3% |
| AA-Omniscience | – | 3.4 |
| GPQA Diamond (AA) | – | 92.7% |
| Humanity's Last Exam (AA) | – | 43.0% |
| SciCode (AA) | – | 53.2% |
| LiveCodeBench | – | 87.8% |
| MMLU-Pro | – | 88.6% |
| IOI | – | 73.0% |
| LegalBench | – | 83.6% |
| CorpFin | – | 65.8% |
| TaxEval | – | 75.5% |
| Terminal-Bench 2.1 (Vals) | 87.3% | 67.4% |
| SWE-bench (Vals) | – | 85.6% |
| GPQA Diamond (Vals) | – | 93.7% |
| Vals Index | 66.6 | 51.8 |
| ARC-AGI-1 | 97.5% | – |
| ARC-AGI-2 | 95.0% | – |
| ARC-AGI-3 | 62.7% | – |
Data as of 2026-09-09. Best configuration of each model; every score links to its source on the model pages.