The AI arena is free today

Open Superagent
AI OrganizationMaintains 5 benchmarks

ARC Prize Foundation: AI benchmarks & research

Non-profit running the ARC-AGI benchmarks and prize.

Maintained by ARC Prize Foundation

5 benchmarks
  • Arc

    The Abstraction and Reasoning Corpus (ARC) is a benchmark designed to measure human-like general fluid intelligence through grid-based reasoning tasks. It consists of 800 tasks (400 training, 400 evaluation) where each task presents input-output grids that require understanding abstract patterns and transformations. Test-takers must produce exactly correct output grids for all test inputs in a task to solve it, with 3 trials allowed per test input. ARC aims to enable fair comparisons of general intelligence between AI systems and humans using priors designed to be as close as possible to innate human priors.

    Models scored
    1
  • ARC-AGI

    The Abstraction and Reasoning Corpus for Artificial General Intelligence (ARC-AGI) is a benchmark designed to test general intelligence and abstract reasoning capabilities through visual grid-based transformation tasks. Each task consists of 2-5 demonstration pairs showing input grids transformed into output grids according to underlying rules, with test-takers required to infer these rules and apply them to novel test inputs. The benchmark uses colored grids (up to 30x30) with 10 discrete colors/symbols, designed to measure human-like general fluid intelligence and skill-acquisition efficiency with minimal prior knowledge.

    Leader
    GPT-6 Astra98.5%
    Models scored
    12
  • ArcAGI2

    ARC-AGI-2 is the second-generation Abstraction and Reasoning Corpus benchmark measuring fluid, general reasoning and abstraction.

    Leader
    Seed 2.1 Pro62.5%
    Models scored
    3
  • ARC-AGI-3

    ARC-AGI-3 is the third-generation Abstraction and Reasoning Corpus benchmark, an interactive-reasoning evaluation designed to measure fluid, novel problem-solving ability that remains far from saturated for frontier models.

    Leader
    GPT-6 Astra99.9%
    Models scored
    6
  • ARC-AGI v2

    ARC-AGI-2 is an upgraded benchmark for measuring abstract reasoning and problem-solving abilities in AI systems through visual grid transformation tasks. It evaluates fluid intelligence via input-output grid pairs (1x1 to 30x30) using colored cells (0-9), requiring models to identify underlying transformation rules from demonstration examples and apply them to test cases. Designed to be easy for humans but challenging for AI, focusing on core cognitive abilities like spatial reasoning, pattern recognition, and compositional generalization.

    Leader
    GPT-6 Astra95.0%
    Models scored
    19

In one line

ARC Prize Foundation maintains Arc (counts toward the LLM Stats Score, 1 model scored), ARC-AGI (counts toward the LLM Stats Score, 12 models scored), ArcAGI2 (counts toward the LLM Stats Score, 3 models scored), ARC-AGI-3 (counts toward the LLM Stats Score, 6 models scored) and ARC-AGI v2 (counts toward the LLM Stats Score, 19 models scored). It ships no models of its own on LLM Stats.

FAQ

Common questions about ARC Prize Foundation.

What is ARC Prize Foundation?

Non-profit running the ARC-AGI benchmarks and prize. On LLM Stats it maintains 5 benchmarks, led by ARC-AGI v2, ARC-AGI and ARC-AGI-3. It ships no models of its own; each score on its benchmarks carries its own provenance label on the model page.

Which ARC Prize Foundation benchmarks count toward the LLM Stats Score?

ARC-AGI v2, ARC-AGI, ARC-AGI-3, ArcAGI2 and Arc count toward the LLM Stats Score. Only benchmarks admitted to the core set move the main leaderboard; the rest are shown for transparency.

Which models lead ARC Prize Foundation's benchmarks?

GPT-6 Astra leads ARC-AGI v2 at 95.0% (19 models scored), GPT-6 Astra leads ARC-AGI at 98.5% (12 models scored) and GPT-6 Astra leads ARC-AGI-3 at 99.9% (6 models scored). Each benchmark page lists the full ranking and how the results were produced.

Where are ARC Prize Foundation's official benchmark results?

ARC Prize Foundation's official site is https://arcprize.org. Each benchmark page on LLM Stats links to that benchmark's own paper or leaderboard, and shows the scores next to every other benchmark so models can be compared on one page.