Unverified50% confidenceFactExact time
When ARC-AGI 3 was released, frontier models scored below 1% on all tasks, while every task is solvable by humans.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/24/2026
First Seen
Valid until: 1/22/2027
Sources
Related Entities
Related Claims
UnverifiedARC-AGI-2 was designed by AI safety researcher François Chollet; early GPT-4 scored less than 10% on this test, while the human average score is approximately 85%68% similarUnverifiedOn ARC-AGI-1, before 2024, the accuracy of state-of-the-art AI systems hovered around 30% for a long time, far below the human level of over 85%.66% similarUnverifiedIn Artificial Analysis' independent evaluation, Argon completes benchmark tasks at only $1.99 per task, close to GLM 5.3 and clearly lower than all frontier competitors except Sol 6.1.66% similarUnverifiedOn ARC-AGI 3, Astra completed 96% of levels at the highest setting using fewer steps than the human baseline, reducing actions by approximately 50% on average.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1092250API
curl https://kongchang.com/api/v1/knowledge/claims/1092250MCP
get_claim(id=1092250)