| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Data Science Agent tasks | xBench-DS | Pass@173 | 31 | |
| Deep Search | XBench-DS v1.0 (test) | Success Rate68 | 22 | |
| Information-seeking | Xbench-DS | Success Rate75 | 18 | |
| Data Science Task Automation | xBench-DS | Score78 | 18 | |
| Web Browsing Research | XBench-DS (XBD) | Pass@389 | 6 |