ARC-AGI-1
Created by François Chollet, 5 Nov 2019 · Higher is better · Website
- OpenAI
- Anthropic
- Reported in a later post or card
Results
| Model | Score | Source |
|---|---|---|
| GPT-6 Astra3 Sep 2026 · OpenAI | 98.5% | GPT-6 Astra: A new generation of intelligenceChecked A person checked this number against its source on 18 Oct 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.Our capture |
| GPT-5.6 Sol9 Jul 2026 · OpenAI | 97.5% | GPT-6 Astra: A new generation of intelligenceChecked A person checked this number against its source on 29 Oct 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.highest score at any effort; reported later |
| Claude Opus 4.65 Feb 2026 · Anthropic | 94.00% | Claude Opus 4.6 System CardCorroborated The same number is printed in 1 other lab document: 1Run by ARC Prize Foundation · high effort; 120K thinking tokens; private set |
| Claude Opus 4.716 Apr 2026 · Anthropic | 93.5% | Claude Opus 4.7 System CardChecked A person checked this number against its source on 28 Oct 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.Run by ARC Prize Foundation · high effort; private set |
| Claude Opus 4.716 Apr 2026 · Anthropic | 92.0% | Claude Opus 4.7 System CardChecked A person checked this number against its source on 28 Oct 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.Run by ARC Prize Foundation · max effort; private set; another report, not charted |
| Claude Sonnet 4.617 Feb 2026 · Anthropic | 86.50% | Claude Sonnet 4.6 System CardChecked A person checked this number against its source on 28 Oct 2026. This confirms the number matches what the lab published. It doesn't confirm that the score was reproduced.Run by ARC Prize Foundation · high effort; 120K thinking tokens; private set |
