Claude Opus 4.7 (Non-reasoning, High): benchmark results
Provider: Anthropic. Released 2026-04-16. Access: API.
Unified ELO 1647 ± 1, rank #253 of 1761 rated models, from 18 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA Terminal-Bench Hard | 54.55 | Accuracy (%) | 96.8 |
| AA Omniscience - Law | 41.8 | Accuracy (%) | 89.9 |
| AA Omniscience | 14.83 | Score | 89.8 |
| AA Omniscience - Business | 38.3 | Accuracy (%) | 89.4 |
| AA Omniscience - Software Engineering (SWE) | 66.6 | Accuracy (%) | 88.1 |
| AA-Omniscience Accuracy | 44.73 | Accuracy (%) | 87.7 |
| Artificial Analysis Intelligence Index | 35.43 | Intelligence Index | 87.4 |
| AA Omniscience - Science, Engineering & Mathematics | 45.18 | Accuracy (%) | 87.3 |
| AA GPQA Diamond | 88.48 | Accuracy (%) | 85.2 |
| AA Omniscience - Health | 38.65 | Accuracy (%) | 85.2 |
| AA Humanity's Last Exam | 33.32 | Accuracy (%) | 84.4 |
| AA Omniscience - Humanities & Social Sciences | 38.9 | Accuracy (%) | 83.6 |
Interactive version: theaggregate.ai/model?slug=claude-opus-4-7-non-reasoning-high · How It Works · Data refreshed daily, snapshot 2026-09-05.