Benchmark
ARC-AGI-2 Benchmark
What is ARC-AGI-2?
Cost per task is part of the official result and is routinely dropped when the score is quoted on its own. Benchmark Radar tracks 1 reported scores for…
- Released
- 2025-03-24
- Evaluates
- reasoning
- Openness
- unknown
- Importer
- Model reports
- Reported scores
- 1
Source provenance
- Original evidence https://arcprize.org/arc-agi/2/
ARC-AGI-2 links
- Website https://arcprize.org/arc-agi/2/
ARC-AGI-2 results and reported scores
Model reports
| Model | Organization | Reported value | Reported | Evidence |
|---|---|---|---|---|
| Gemini 3.1 Pro | 77.1 | 2026-02-19 | evidence |
Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.
Which sources cite ARC-AGI-2?
- Grok 4.5 xAI, 2026-07-08
- Gemini 3.1 Pro Google, 2026-02-19