| code-generation-on-programbench-golden-166-no-time-limit | Claude Opus 5.5 (max effort) (mini-swe-agent) | #1 | Average Test Pass Rate: 91.2 |
| code-generation-on-swe-bench-multilingual | Claude Opus 5.5 | #1 | Accuracy: 93.9 |
| code-generation-on-swe-bench-multimodal | Claude Opus 5.5 (max effort) | #1 | Accuracy: 61.4 |
| code-generation-on-swe-bench-pro | Claude Opus 5.5 | #1 | Resolved: 89.9 |
| computer-use-agents-on-osworld-2-1-108-tasks | Claude Opus 5.5 (max effort) | #1 | Partial: 81.8Binary: 48.7 |
| multi-task-language-understanding-on-global-mmlu | Claude Opus 5.5 (max effort) | #1 | Accuracy: 94.3 |
| multi-task-language-understanding-on-milu | Claude Opus 5.5 (max effort) | #1 | Accuracy: 93.1 |
| question-answering-on-officeqa-pro-preselected-text | Claude Opus 5.5 (max effort) | #1 | Accuracy: 67.7 |
| chart-question-answering-on-chartography | Claude Opus 5.5 (with tools, max effort) | #2 | Pass@1: 0.89 |
| code-generation-on-terminal-bench-4-0 | Claude Opus 5.5 (Claude Code) | #2 | Accuracy: 66.36 |
| medical-question-answering-on-healthbench-opus-4-8-grader-length-adjusted | Claude Opus 5.5 (max effort, Opus 5 refusal fallback) | #2 | Score: 60.6 |
| medical-question-answering-on-healthbench-professional | Claude Opus 5.5 | #2 | Score: 65.6 |
| agents-on-terminal-bench-science-0-1 | Claude Opus 5.5 (Claude Code) | #3 | Accuracy: 58.7 |
| agents-on-toolathlon-verified-108-tasks | Claude Opus 5.5 | #3 | Pass@1: 0.778 |
| code-generation-on-deepswe | Claude Opus 5.5 | #4 | Pass@1: 0.742 |