← Benchmarks
CursorBench v3.2
CursorBench v3.2 evaluates coding agents on interactive software engineering tasks in the Cursor environment.
id cursorbench-3.2 · max 1 · 2 models reported
| # | Model | Score |
|---|---|---|
| 1 | Claude Fable 5.1 Anthropic | 0.73 |