# ARC-AGI-3

ARC-AGI-3 is an AI benchmark; this page records what it measures and traces its data, grader, and scores to primary sources.

Tests exploration, goal discovery, world-model building, and planning in novel interactive environments.

## Details

- Address: Online, Global
- Website: [arcprize.org/arc-agi/3](<https://arcprize.org/arc-agi/3>)
- Tags: Viral, Science
- Category: Agents
- Status: Alpha
- Data licence: unknown
- Data access: open
- Grader: simulation
- Organization: ARC Prize Foundation
- Released: 2026-03
- Paper: [arxiv.org/abs/2603.24621](<https://arxiv.org/abs/2603.24621>)
- Leaderboard: [arcprize.org/arc-agi/3/leaderboard](<https://arcprize.org/arc-agi/3/leaderboard>)
- Top-score provenance: unknown
- ABI family: No

## Questions

### Where is ARC-AGI-3?

ARC-AGI-3 is at Online, Global.

### How do I contact ARC-AGI-3?

To contact ARC-AGI-3, visit https://arcprize.org/arc-agi/3.

## Source

From [Benchmark primary sources](https://aibenchmarkindex.com/about), license CC0-1.0, checked 2026-10-15.

Last updated October 15, 2026.

## Links

- This benchmark on the [web page](<https://aibenchmarkindex.com/global/online/arc-agi-3>)
- [Agents in Online](<https://aibenchmarkindex.com/global/online/category/agents.md>)
- [More benchmarks in Online](<https://aibenchmarkindex.com/global/online.md>)

> **Check before you compare.** Benchmark methods and leaderboards change. Follow the cited primary source before using a result in a decision.