{
  "id": "arc-agi",
  "code": "PTL-0192",
  "term": "ARC-AGI",
  "aliases": [
    "ARC",
    "Abstraction and Reasoning Corpus",
    "ARC-AGI-2",
    "ARC Prize"
  ],
  "category": "benchmarks",
  "definition": "ARC-AGI (the Abstraction and Reasoning Corpus) is a benchmark of visual grid puzzles in which a model must infer a transformation rule from a few examples and apply it to a new grid, designed to measure how efficiently a system acquires new skills.",
  "description": "The puzzles are easy for most people but require reasoning that cannot be memorized, because each task is novel. François Chollet introduced it in 2019 alongside a definition of intelligence as skill-acquisition efficiency. Successive versions, such as ARC-AGI-2, were made harder as systems improved, and many results are reported with explicit compute costs.",
  "example": null,
  "broader": [
    "benchmark"
  ],
  "narrower": [],
  "related": [
    "test-time-compute-scaling",
    "reasoning-model"
  ],
  "introduced": 2019,
  "added": "2026-10-10",
  "sources": [
    {
      "title": "On the Measure of Intelligence",
      "authors": "Chollet",
      "year": 2019,
      "url": "https://arxiv.org/abs/1911.01547"
    },
    {
      "title": "ARC-AGI-2: A New Challenge for Frontier AI Reasoning Systems",
      "authors": "Chollet et al.",
      "year": 2025,
      "url": "https://arxiv.org/abs/2505.11831"
    }
  ],
  "url": "https://protologue.com/t/arc-agi/",
  "citation": "Protologue. (2026). ARC-AGI. In Protologue: A Taxonomy of Prompting and LLM Techniques (v1.4.0, PTL-0192). https://protologue.com/t/arc-agi/"
}