- AA-Omniscience Accuracy22%#106
- Humanity's Last Exam5.2%#204
- AA-LCR64.7%#118
- CritPt0%#143
- MMMU-Pro62.7%#82
- GPQA Diamond75.1%#94
- IFBench79.6%#4
- τ²-Bench Telecom90.6%#12
- Terminal-Bench Hard17.4%#42
Nova 2.0 Pro Preview (Low)Median
Data from Artificial Analysis