Benchmarks
347 sources · 1,258 models × 3,038 benchmarks · 158,912 scoresConcept-learning benchmark for abstraction and program-like reasoning from input-output examples.
No signal or difficulty estimate. No usable scores.
This is a benchmark-family page. Scores are recorded on the component benchmarks below rather than directly on the parent, because the components use different score scales and directions.
| Metric | Models | Leader | Direction | Date |
|---|---|---|---|---|
| ARC-AGI-1 Semi-Private | 93 | Claude Fable 5.298.5 | higher is better | 2025-11-24 |