Claude Opus 4.7 vs GPT-5.2
Claude Opus 4.7 
GPT-5.2 Verdict
- Claude Opus 4.7 leads on quality: 65.6 vs 61.6.
- Claude Opus 4.7 is stronger in agents & tools, coding, composite, human preference, knowledge, maths, multimodal, reasoning.
- GPT-5.2 is stronger in instruction following, long context.
- GPT-5.2 is 2.1× cheaper ($4.81 vs $10.00 per 1M blended).
- GPT-5.2 streams 1.6× faster (68 vs 43 tokens per second).
| Metric | Claude Opus 4.7 | GPT-5.2 |
|---|---|---|
| BenchLeader Index | 65.6 | 61.6 |
| Agents & tools score | 69.4 | 61.2 |
| Coding score | 64.8 | 53.7 |
| Composite score | 80.4 | 67.5 |
| Human preference score | 69.1 | 67.9 |
| Instruction following score | 58.7 | 67.4 |
| Knowledge score | 66.5 | 61.1 |
| Long context score | 65.9 | 67.9 |
| Maths score | 65.4 | – |
| Multimodal score | 65.7 | 59.9 |
| Reasoning score | 66.6 | 61.3 |
| Blended price $/M | $10.00 | $4.81 |
| Output speed | 43 tok/s | 68 tok/s |
| Time to first answer | 21.7 s | 129.4 s |
| Context window | 1M | 400k |
| Humanity's Last Exam | 36.2% | 27.8% |
| Terminal-Bench | 80.2% | 64.9% |
| SimpleBench | 61.7% | 45.8% |
| Remote Labor Index | – | 2.1% |
| WeirdML | 76.4% | – |
| APEX-Agents | – | 23.0% |
| FrontierCode | 38.5% | – |
| GSO-Bench | 44.1% | – |
| Epoch Capabilities Index | 156.3 | 153.5 |
| LMArena Text | 1495 | 1476 |
| LMArena Hard Prompts | 1519 | 1497 |
| LMArena Coding | 1547 | 1515 |
| LMArena WebDev | 1557 | 1417 |
| LMArena Vision | 1317 | 1268 |
| AA Intelligence Index | 40.7 | 30.4 |
| IFBench | 58.6% | 75.4% |
| AA-LCR | 78.7% | 82.7% |
| MMMU-Pro | 78.8% | – |
| AA-Omniscience | 27.3 | -0.9 |
| Terminal-Bench Hard | 51.5% | 47.0% |
| GPQA Diamond (AA) | 91.4% | 90.3% |
| Humanity's Last Exam (AA) | 42.3% | 37.7% |
| τ²-Bench Telecom (AA) | 88.6% | 84.8% |
| AIME (Vals) | 96.3% | – |
| LiveCodeBench | 85.1% | – |
| MMLU-Pro | 89.9% | – |
| IOI | 47.1% | – |
| 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 | – |
| SWE-Bench Pro | – | 29.9% |
| VISTA | – | 46.6% |
| MultiNRC | – | 42.2% |
| HiL-Bench | 41.7% | – |
| TutorBench | – | 53.5% |
| EQ-Bench 4 | 1311 | – |
| Kagi LLM Benchmark | 73.3% | 73.3% |
| SWE-bench Verified (bash only) | – | 69.0% |
| ARC-AGI-1 | – | 94.5% |
| ARC-AGI-2 | – | 72.9% |
| BFCL Overall | – | 55.9% |
Data as of 2026-09-10. Best configuration of each model; every score links to its source on the model pages.
Claude Opus 4.7 vs GPT-5.2: questions
- Is Claude Opus 4.7 better than GPT-5.2?
- Claude Opus 4.7 leads on quality: 65.6 vs 61.6. The BenchLeader Index combines every independent quality benchmark; Claude Opus 4.7 is ahead overall as of 2026-09-10, but check the category scores for your use.
- Is Claude Opus 4.7 better than GPT-5.2 for coding?
- Claude Opus 4.7 scores higher in coding (65 vs 54 on the category index, where 50 is average).
- Is Claude Opus 4.7 better than GPT-5.2 for agentic tasks?
- Claude Opus 4.7 scores higher in agentic tasks (69 vs 61 on the category index, where 50 is average).
- Which is cheaper, Claude Opus 4.7 or GPT-5.2?
- GPT-5.2 is cheaper: $4.81 against $10.00 per million tokens, blended at three input tokens per output token.
- Which is faster, Claude Opus 4.7 or GPT-5.2?
- GPT-5.2 streams faster: 68 against 43 output tokens per second.
- Which has the larger context window?
- Claude Opus 4.7 accepts more context: 1M against 400k tokens.