Claude Opus 4.7 vs Qwen3.8 2.4T A95B
Claude Opus 4.7 
Qwen3.8 2.4T A95B Verdict
- Claude Opus 4.7 and Qwen3.8 2.4T A95B are level on quality (64.5 vs 65.2).
- Claude Opus 4.7 is stronger in coding, human preference, maths, multimodal.
- Qwen3.8 2.4T A95B is stronger in agents & tools, knowledge, reasoning, composite, long context.
- Qwen3.8 2.4T A95B is 3.3× cheaper ($3.00 vs $10.00 per 1M blended).
- Claude Opus 4.7 streams 1.4× faster (53 vs 38 tokens per second).
| Metric | Claude Opus 4.7 | Qwen3.8 2.4T A95B |
|---|---|---|
| BenchLeader Index | 64.5 | 65.2 |
| Agents & tools score | 67.5 | 78.3 |
| Coding score | 64.7 | – |
| Human preference score | 69.0 | – |
| Knowledge score | 62.3 | 66.3 |
| Maths score | 67.5 | – |
| Multimodal score | 67.7 | – |
| Reasoning score | 66.5 | 80.5 |
| Composite score | – | 79.8 |
| Long context score | – | 66.6 |
| Blended price $/M | $10.00 | $3.00 |
| Output speed | 53 tok/s | 38 tok/s |
| Time to first answer | 1.6 s | 55.2 s |
| Context window | 1M | 984k |
| Humanity's Last Exam | 36.2% | – |
| Terminal-Bench | 80.2% | – |
| SimpleBench | 61.7% | – |
| WeirdML | 76.4% | – |
| FrontierCode | 38.5% | – |
| GSO-Bench | 44.1% | – |
| Epoch Capabilities Index | 156.3 | – |
| LMArena Text | 1494 | – |
| LMArena Hard Prompts | 1519 | – |
| LMArena Coding | 1547 | – |
| LMArena WebDev | 1557 | – |
| LMArena Vision | 1318 | – |
| AA Intelligence Index | – | 40.0 |
| AA-LCR | – | 80.3% |
| AA-Omniscience | – | 4.3 |
| GPQA Diamond (AA) | – | 93.5% |
| Humanity's Last Exam (AA) | – | 42.5% |
| SciCode (AA) | – | 54.0% |
| AIME (Vals) | 96.3% | – |
| LiveCodeBench | 85.1% | – |
| MMLU-Pro | 89.9% | – |
| LegalBench | 85.3% | – |
| CorpFin | 66.1% | – |
| TaxEval | 75.3% | – |
| Terminal-Bench 2.1 (Vals) | 68.5% | – |
| SWE-bench (Vals) | 82.0% | – |
| GPQA Diamond (Vals) | 90.2% | – |
| Vals Index | 56.1 | – |
| HiL-Bench | 41.7% | – |
| EQ-Bench 4 | 1311 | – |
| Kagi LLM Benchmark | 73.3% | – |
| CritPt | – | 20.0% |
| GDPval (AA) | – | 56.4% |
| τ²-Bench Banking (AA) | – | 49.1% |
| CaseLaw v2 | 68.4% | – |
| Code Migration | 43.9% | – |
| CyberBench | 52.1% | – |
| Excel Modeling Benchmark | 63.7% | – |
| Finance Agent v2 | 51.5% | – |
| Harvey's Legal Agent Benchmark | 6.7% | – |
| Legal Research Bench | 38.5% | – |
| MedCode | 54.9% | – |
| MedScribe | 83.0% | – |
| MMMU-Pro (Vals) | 85.5% | – |
| MortgageTax | 70.3% | – |
| ProgramBench | 0.0% | – |
| SAGE | 56.1% | – |
| Terminal-Bench 2.0 (Vals) | 68.5% | – |
| Vals Multimodal Index | 67.4% | – |
| Vibe Code Bench v1.1 | 71.0% | – |
| SWE Atlas: Codebase QnA | 40.3% | – |
| SWE Atlas: Refactoring | 48.6% | – |
| SWE Atlas: Test Writing | 38.5% | – |
| LMArena Maths | 1491 | – |
| LMArena Creative Writing | 1481 | – |
| LMArena Instruction Following | 1493 | – |
| LMArena Multi-turn | 1515 | – |
| LMArena Longer Queries | 1507 | – |
| LMArena Search | 1212 | – |
| LMArena Document | 1494 | – |
| Mystery Game Puzzles | 13.0% | – |
| ExploitBench | 26.5% | – |
| ForecastBench | 59.6% | – |
| GBAEval | 43.8% | – |
| ALE-Bench | 1323.0 | – |
| Vending-Bench 2 | 10936.8 | – |
| Blueprint-Bench 2 | 24.5% | – |
Data as of 2026-09-19. Best configuration of each model; every score links to its source on the model pages.
Claude Opus 4.7 vs Qwen3.8 2.4T A95B: questions
- Is Claude Opus 4.7 better than Qwen3.8 2.4T A95B?
- Claude Opus 4.7 and Qwen3.8 2.4T A95B are level on quality (64.5 vs 65.2). The BenchLeader Index combines every independent quality benchmark; Qwen3.8 2.4T A95B is ahead overall as of 2026-09-19, but check the category scores for your use.
- Is Claude Opus 4.7 better than Qwen3.8 2.4T A95B for agentic tasks?
- Qwen3.8 2.4T A95B scores higher in agentic tasks (78 vs 68 on the category index, where 50 is average).
- Which is cheaper, Claude Opus 4.7 or Qwen3.8 2.4T A95B?
- Qwen3.8 2.4T A95B is cheaper: $3.00 against $10.00 per million tokens, blended at three input tokens per output token.
- Which is faster, Claude Opus 4.7 or Qwen3.8 2.4T A95B?
- Claude Opus 4.7 streams faster: 53 against 38 output tokens per second.
- Which has the larger context window?
- Claude Opus 4.7 accepts more context: 1M against 984k tokens.