Collaborative Execution on Agentization Bench Hard
44.4Original Success Rate (Or.SR)Claude Code
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Claude Code2026.04 | 44.4 | 7.4 | |
| EnvX2026.04 | 44.4 | 11.1 | |
| OpenHands2026.04 | 29.6 | 11.1 | |
| Codex2026.04 | 25.9 | 7.4 |
| Method | Links | ||
|---|---|---|---|
| Claude Code2026.04 | 44.4 | 7.4 | |
| EnvX2026.04 | 44.4 | 11.1 | |
| OpenHands2026.04 | 29.6 | 11.1 | |
| Codex2026.04 | 25.9 | 7.4 |