ARC Prize Foundation GTM brain

ARC Prize Foundation develops AGI benchmarks and prizes for AI researchers, labs and the open research community.

category

AI capability benchmarks and research prizes

market profile

  • text: The nonprofit frames general intelligence as efficient adaptation to novel problems and argues that scaling alone is insufficient.
  • text: It combines benchmark releases, leaderboards, open research incentives and competitions; it identifies NIST CAISI as a benchmark recipient.

personas

  • title: Frontier AI evaluation lead
    pains:
    • Common benchmarks can reward memorization instead of measuring adaptation to new tasks.
    • Model teams need credible comparisons of reasoning and agentic progress.
  • title: Independent AI researcher or competition participant
    pains:
    • Novel research approaches need public evaluation and resources.
    • Closed research practices can limit reproducibility and access.
  • title: AI research funder or policy evaluator
    pains:
    • Capability claims are difficult to compare without rigorous, human-calibrated measures.
    • Funders need ways to direct attention toward meaningful research bottlenecks.

icp

  • text: Frontier AI labs and academic research groups evaluating generalization, reasoning and agentic systems.
  • text: Independent open-source researchers, competition participants and organizations funding AI research or standards.

message pillars

  • text: Advance open research with benchmarks and prizes.
  • text: Measure capabilities that remain challenging for AI despite being easy for people.
  • text: Reward research that makes progress openly available.

objections

  • objection: Does a single benchmark fully measure general intelligence?
    response: Use ARC-AGI alongside complementary evaluations and inspect task design and test policy before interpreting a score.
  • objection: Can public tasks be memorized or contaminated?
    response: Review benchmark versions, testing policy and evaluation methodology when comparing model claims.
  • objection: Why support another benchmark?
    response: The foundation combines human-calibrated tasks with prizes and open research incentives to surface new approaches.

competitors

  • name: Epoch AI
    website: https://epoch.ai/
    why: Publishes AI capabilities research and alternative evaluations including FrontierMath.
  • name: Artificial Analysis
    website: https://artificialanalysis.ai/
    why: Offers independent model evaluation, an intelligence index and leaderboards.

influencers

  • name: François Chollet
    url: https://arcprize.org/about
    why: Co-founder and creator of ARC-AGI.
  • name: Mike Knoop
    url: https://arcprize.org/about
    why: Co-founder and board member.

keywords

  • ARC AGI benchmark
  • general intelligence benchmark
  • AI reasoning evaluation
  • fluid intelligence AI
  • agentic AI benchmark
  • open AGI research competition

voice

  • Mission-driven and open research oriented
  • Confident and scientifically framed

signals

  • text: An AI lab announces a new reasoning or agentic model and needs independent evaluation.
  • text: A research funder or institution seeks open, human-calibrated benchmarks for generalization.

pricing

model: unknown

motion

community_led

channels

  • competitions
  • leaderboards
  • research content
  • Discord
  • X
  • YouTube
  • GitHub
  • events

confidence

medium

Founders

  • name: Greg Kamradt
    title: Founder/President
    role: Founder
    bio: Greg is the president of ARC Prize Foundation. He leads the nonprofit's work building benchmarks that have been used by OpenAI, Anthropic, Google, and xAI. Previously, Greg led ops + product at Digits (AI Accounting), and spent several years at Salesforce driving strategy, growth, and predictive analytics. Greg is also the founder of Leverage, teaching 100K+ developers how to build their first AI applications through his YouTube channel (66K subs).

Back to YC companies