Claude Opus 5 vs Qwen3 8
Claude Opus 5 
Qwen3 8 Verdict
- Claude Opus 5 leads on quality: 70.0 vs 63.9.
- Claude Opus 5 is stronger in agents & tools, coding, composite, human preference, knowledge, long context, maths, multimodal, reasoning.
- Qwen3 8 (max) is 3.8× cheaper ($2.67 vs $10.00 per 1M blended).
- Claude Opus 5 streams 1.3× faster (52 vs 39 tokens per second).
| Metric | Claude Opus 5 | Qwen3 8 (max) |
|---|---|---|
| BenchLeader Index | 70.0 | 63.9 |
| Agents & tools score | 71.6 | 56.1 |
| Coding score | 74.4 | 71.7 |
| Composite score | 92.9 | 70.8 |
| Human preference score | 78.2 | 67.9 |
| Knowledge score | 70.5 | 62.3 |
| Long context score | 66.2 | 65.7 |
| Maths score | 67.1 | – |
| Multimodal score | 69.4 | 67.2 |
| Reasoning score | 71.1 | 67.7 |
| Blended price $/M | $10.00 | $2.67 |
| Output speed | 52 tok/s | 39 tok/s |
| Time to first answer | 81.6 s | 53.2 s |
| Context window | 1M | 1M |
| GPQA Diamond | 92.9% | – |
| OTIS Mock AIME | 97.8% | – |
| SimpleBench | 80.6% | – |
| WeirdML | 86.3% | – |
| Epoch Capabilities Index | 162.6 | 156.6 |
| LMArena Text | – | 1480 |
| LMArena Hard Prompts | – | 1502 |
| LMArena Coding | – | 1520 |
| LMArena WebDev | – | 1685 |
| LMArena Vision | – | 1313 |
| LMArena Agent | – | 3.9 |
| LiveBench | – | 78.5% |
| LiveBench Reasoning | – | 88.2% |
| LiveBench Coding | – | 72.9% |
| LiveBench Agentic Coding | – | 64.7% |
| LiveBench Mathematics | – | 91.3% |
| LiveBench Data Analysis | – | 78.4% |
| LiveBench Language | – | 79.7% |
| AA Intelligence Index | 50.7 | 40.3 |
| AA-LCR | 79.3% | 78.3% |
| MMMU-Pro | 84.7% | 82.3% |
| AA-Omniscience | 37.1 | 3.4 |
| GPQA Diamond (AA) | 93.2% | 92.7% |
| Humanity's Last Exam (AA) | 54.9% | 43.0% |
| SciCode (AA) | 56.4% | 53.2% |
| LiveCodeBench | 89.0% | 87.8% |
| MMLU-Pro | 91.6% | 88.6% |
| IOI | 91.7% | 73.0% |
| LegalBench | 87.0% | 83.6% |
| CorpFin | 73.2% | 65.8% |
| TaxEval | 75.1% | 75.5% |
| Terminal-Bench 2.1 (Vals) | 84.6% | 67.4% |
| SWE-bench (Vals) | 97.0% | 85.6% |
| GPQA Diamond (Vals) | 93.4% | 93.7% |
| Vals Index | 67.2 | 51.8 |
| HiL-Bench | 57.0% | – |
| EQ-Bench 4 | 1385 | – |
Data as of 2026-09-09. Best configuration of each model; every score links to its source on the model pages.