Claude Opus 4.8 is Anthropic's most capable generally available model in the Opus family. It supports text, image, and file inputs with text output, with reasoning support and a 1M-token...
| Benchmark | Score | Source |
|---|---|---|
| SWE-bench Verified | 88.6 | verified |
| SWE-Bench Pro | 69.2 | verified |
| Benchmark | Score | Source |
|---|---|---|
| GPQA Diamond | 93.6 | verified |
| Humanity's Last Exam |
57.9 |
| verified |
| Benchmark | Score | Source |
|---|---|---|
| CharXiv Reasoning | 89.9 | verified |
| ScreenSpot Pro | 87.9 | verified |
| Benchmark | Score | Source |
|---|---|---|
| BrowseComp | 84.3 | verified |
| MCP Atlas | 82.2 | verified |
| Toolathlon | 59.9 | verified |
Anthropic