Claude Opus 4.5 vs Claude Opus 4.6
Claude Opus 4.5 Verdict
- Claude Opus 4.6 leads on quality: 63.3 vs 60.7.
- Claude Opus 4.5 (thinking) is stronger in agents & tools, instruction following.
- Claude Opus 4.6 is stronger in coding, composite, knowledge, long context, maths, multimodal, reasoning, human preference.
- They cost about the same ($10.00 per 1M blended).
| Metric | Claude Opus 4.5 (thinking) | Claude Opus 4.6 |
|---|---|---|
| BenchLeader Index | 60.7 | 63.3 |
| Agents & tools score | 74.6 | 68.8 |
| Coding score | 56.4 | 62.4 |
| Composite score | 65.8 | 69.4 |
| Instruction following score | 54.8 | 53.9 |
| Knowledge score | 61.0 | 70.6 |
| Long context score | 65.2 | 65.5 |
| Maths score | 64.0 | 65.7 |
| Multimodal score | 58.0 | 63.6 |
| Reasoning score | 58.6 | 62.6 |
| Human preference score | – | 64.9 |
| Blended price $/M | $10.00 | $10.00 |
| Output speed | 45 tok/s | 36 tok/s |
| Time to first answer | 17.3 s | 2.4 s |
| Context window | 200k | 1M |
| GPQA Diamond | – | 90.5% |
| OTIS Mock AIME | – | 94.4% |
| SWE-bench Verified (Epoch) | – | 78.7% |
| Humanity's Last Exam | – | 19.0% |
| Terminal-Bench | – | 79.8% |
| SimpleBench | – | 67.6% |
| Cybench | – | 93.0% |
| Remote Labor Index | – | 4.2% |
| WeirdML | – | 77.9% |
| APEX-Agents | – | 32.4% |
| FrontierCode | – | 26.6% |
| GSO-Bench | – | 33.3% |
| Epoch Capabilities Index | – | 155.3 |
| LMArena Text | – | 1498 |
| LMArena Hard Prompts | – | 1527 |
| LMArena Coding | – | 1546 |
| LMArena WebDev | – | 1537 |
| LMArena Vision | – | 1311 |
| AA Intelligence Index | 29.1 | 31.9 |
| IFBench | 58.0% | 53.1% |
| AA-LCR | 77.3% | 78.0% |
| MMMU-Pro | 74.0% | 75.4% |
| AA-Omniscience | 14 | 13.7 |
| Terminal-Bench Hard | 47.0% | 48.5% |
| GPQA Diamond (AA) | 86.6% | 89.6% |
| Humanity's Last Exam (AA) | 30.1% | 39.9% |
| τ²-Bench Telecom (AA) | 89.5% | 92.1% |
| AIME (Vals) | 95.4% | – |
| LiveCodeBench | 83.7% | – |
| MMLU-Pro | 87.3% | – |
| IOI | 20.3% | – |
| LegalBench | 84.6% | – |
| CorpFin | 65.1% | – |
| TaxEval | 74.9% | – |
| MedQA | 95.9% | – |
| MGSM | 95.2% | – |
| SWE-bench (Vals) | 76.4% | – |
| GPQA Diamond (Vals) | 85.9% | – |
| MultiChallenge | 59.0% | – |
| PRBench Finance | 46.2% | – |
| PRBench Legal | 44.2% | – |
| VISTA | 46.4% | – |
| MultiNRC | 48.6% | – |
| HiL-Bench | – | 38.3% |
| TutorBench | 51.2% | – |
| EQ-Bench 4 | – | 1223 |
| Kagi LLM Benchmark | 80.2% | 72.4% |
| SWE-bench Verified (bash only) | – | 75.6% |
| SWE-bench Verified (any scaffold) | – | 75.6% |
| ARC-AGI-1 | 80.0% | – |
| ARC-AGI-2 | 37.6% | – |
Data as of 2026-09-10. Best configuration of each model; every score links to its source on the model pages.
Claude Opus 4.5 vs Claude Opus 4.6: questions
- Is Claude Opus 4.5 better than Claude Opus 4.6?
- Claude Opus 4.6 leads on quality: 63.3 vs 60.7. The BenchLeader Index combines every independent quality benchmark; Claude Opus 4.6 is ahead overall as of 2026-09-10, but check the category scores for your use.
- Is Claude Opus 4.5 better than Claude Opus 4.6 for coding?
- Claude Opus 4.6 scores higher in coding (62 vs 56 on the category index, where 50 is average).
- Is Claude Opus 4.5 better than Claude Opus 4.6 for agentic tasks?
- Claude Opus 4.5 scores higher in agentic tasks (75 vs 69 on the category index, where 50 is average).
- Which is cheaper, Claude Opus 4.5 or Claude Opus 4.6?
- Claude Opus 4.5 is cheaper: $10.00 against $10.00 per million tokens, blended at three input tokens per output token.
- Which is faster, Claude Opus 4.5 or Claude Opus 4.6?
- Claude Opus 4.5 streams faster: 45 against 36 output tokens per second.
- Which has the larger context window?
- Claude Opus 4.6 accepts more context: 1M against 200k tokens.