ProductUnclaimedAIBenchmarksIndustrialsOpen SourceResearch

ARC-AGI Benchmark Series

by ARC Prize Foundation

Open-source AI benchmarks that measure general fluid intelligence and guide AGI research.

arcprize.org

Community-submitted · content unverified. This organization has not verified ownership or the information in this profile.

About

The ARC Prize Foundation publishes the ARC-AGI benchmark series—scientifically grounded evaluations designed to reveal the gap between tasks that are easy for humans but hard for AI. The series includes ARC-AGI-3, described as the world's only unbeaten benchmark measuring agentic intelligence. The benchmarks are used by OpenAI, Anthropic, Google DeepMind, and xAI, and are recognized by NIST's Center for AI Standards and Innovation. Competitions run in partnership with Kaggle, offering over $2M in prizes for open-sourced progress on the benchmarks. ARC-AGI Benchmark Series is a product of ARC Prize Foundation.

Added via
web
Ownership
unclaimed

Where to go

For people

For agents

Nothing listed yet.

Product details

Product kind
Open Source
Pricing model
Free
Free tier
Yes
Open source
Yes

Platforms

web

Use cases

Evaluate AI systems for general fluid intelligenceIdentify inflection points in AI progressGuide open AGI researchRegulatory standards for AI capability measurementCompete for prize funding via Kaggle

Integrations

Kaggle

← Directory