HyperAI
Command Palette
Search for a command to run...
GDP.pdf
An implementation of Surge AI's GDP.pdf benchmark, testing whether language models can reason over long, real-world professional documents and satisfy detailed task-specific criteria.
- Intelligence Index weight
- 5%
- Models
- 148 current, 254 in total
- Median
- 13.2%
- Range
- 0.0%–32.6%