Benchmarks
Every benchmark the index tracks, by tier, with each core benchmark's share of the intelligence index and how many models it covers.
571 benchmarks tracked; 10 weigh in the intelligence index.
- AIME (annual) · 175 models
- ARC-AGI-2 · 68 models
- LMArena / Arena Elo · 167 models
- GPQA Diamond · 288 models
- Humanity's Last Exam · 252 models
- LiveBench · 81 models
- MMLU-Pro · 173 models
- SWE-bench Verified · 110 models
- tau-bench / tau2-bench · 192 models
- Terminal-Bench (Hard) · 232 models