- AA-Omniscience Accuracy18.2%#136
- Humanity's Last Exam9%#163
- AA-LCR60%#131
- CritPt0%#143
- MMMU-Pro62.5%#83
- GPQA Diamond76.8%#81
- IFBench68.5%#22
- τ²-Bench Telecom75.7%#45
- Terminal-Bench Hard17.4%#42
Nova 2.0 Lite (Medium)Median
Data from Artificial Analysis