Claude Opus 4.7 (xHigh): benchmark results
Claude Opus 4.7 evaluated at the xhigh reasoning-effort setting. Provider: Anthropic. Released 2026-04-16. Access: API.
Aggregate Rating 1834 ± 12, rank #86 of 2168 rated models, from 146 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| HealthAgentBench - EHR Event Modelling | 78 | Longitudinal EHR clinical event prediction adapted from EHRS | 100 |
| Nejumi 4 - jaster (0-shot) - CommonsenseMoralJA | 99 | Exact match (%) | 100 |
| RoadmapBench - Completion Score | 69.2 | Completion score (%): mean over tasks of the complexity-weig | 100 |
| RoadmapBench - Infra and Tooling | 43.5 | Resolved rate (%) on the 23 infrastructure and tooling tasks | 100 |
| RoadmapBench - ML and Data | 30.6 | Resolved rate (%) on the 36 machine learning and data tasks | 100 |
| RoadmapBench - UI and Rendering | 64.3 | Resolved rate (%) on the 14 UI and rendering tasks (every su | 100 |
| RoadmapBench - Web and RPC | 41.2 | Resolved rate (%) on the 17 web and RPC tasks (every subtask | 100 |
| SWE Atlas (All Workflows) | 38.94 | Pass@1 (%): mean per-trial pass rate over all 284 SWE Atlas | 100 |
| SWE-Chain | 68.5 | F1-score (%; 2TP / (2TP + FP + FN) over each chain's release | 100 |
| TensorBench | 64.8 | Pass rate (%) on all 199 feature-addition and refactoring ta | 100 |
| TensorBench - Codegen | 71.4 | Pass rate (%) on the 7 code-generation tasks (OpenMP and ato | 100 |
| TensorBench - IR | 62.5 | Pass rate (%) on the 8 intermediate-representation tasks (ne | 100 |
Interactive version: theaggregate.ai/model?slug=claude-opus-4-7-xhigh · How It Works · Data refreshed daily, snapshot 2026-10-20.