To ship ARC-AGI-3, we've set up an entire game studio. Exceptional crew — super proud of the team. They did an incredible job.
@fchollet
-
ARC-AGI-3 Leaderboard: Measuring True AI Intelligence Beyond Human Design
By
–
It is trivial to solve all public ARC-AGI-3 tasks if you have a human looking at them and designing a system to beat them (we have released a harness that uses human replay to score 100%). But our leaderboard is not about measuring how well human intelligence does on ARC-AGI-3,
-
ARC-AGI-3 Benchmarks Agent Performance Against Human Action Efficiency
By
–
ARC-AGI-3 scores agents on how close they are to human action efficiency. All ARC-AGI-3 environments were solved by at least 2 human testers out of 10 (most of the time it was 5+). We use the action count of the 2nd best tester (to avoid outlier performance) as our human
-
AGI Capability: Task Learning Without Human Intervention Required
By
–
Either you believe AGI is possible, in which case a real AGI will be able to look at ARC-AGI-3 and ace it, because regular humans can… …or you believe that AI is just an automation tool that will require human intervention every time a new task comes up. Pick your camp.
-
What General Intelligence Really Means for AGI
By
–
The G in AGI stands for "general". General intelligence does not mean that you have been specifically trained for a large range of tasks. It means you can approach any NEW task and figure it out, just like humans do. If regular people can do it on their own (no guidance, no
-
ARC-AGI-3 Benchmark Monitors Frontier Models AGI Breakthrough
By
–
At the moment, ARC-AGI-3 is the only unsaturated agentic AI benchmark. Sub-1% scores from frontier models on the private test set.
— François Chollet (@fchollet) 25 mars 2026
If you want to be among the first to know when an AGI breakthrough happens, monitor the ARC-AGI-3 leaderboard. Any sudden score jump will mean… https://t.co/EenOUgpNnwAt the moment, ARC-AGI-3 is the only unsaturated agentic AI benchmark. Sub-1% scores from frontier models on the private test set. If you want to be among the first to know when an AGI breakthrough happens, monitor the ARC-AGI-3 leaderboard. Any sudden score jump will mean
-
ARC-AGI-2 Kaggle Competition Final Round With Unlimited Prize
By
–
We're also running one last ARC-AGI-2 competition on Kaggle this year. Get your high score in: since this is the last official ARC-AGI-2 competition, the grand prize will go to the top score regardless of whether it's above the 85% threshold.
-
ARC-AGI-3 Kaggle Competition Tests AI Agents
By
–
You can also enter the ARC-AGI-3 competition on Kaggle. Your AI agents will be tested on two separate private test sets of 55 environments.
-
ARC-AGI-3 Benchmark Evaluates Agentic Intelligence Systems
By
–
ARC-AGI-3 is out now! We've designed the benchmark to evaluate agentic intelligence via interactive reasoning environments. Beating ARC-AGI-3 will be achieved when an AI system matches or exceeds human-level action efficiency on all environments, upon seeing them for the first… pic.twitter.com/zHLOS1ncr7
— François Chollet (@fchollet) 25 mars 2026ARC-AGI-3 is out now! We've designed the benchmark to evaluate agentic intelligence via interactive reasoning environments. Beating ARC-AGI-3 will be achieved when an AI system matches or exceeds human-level action efficiency on all environments, upon seeing them for the first
-
High-Fluid Intelligence Systems Will Dominate Knowledge-Dependent AI
By
–
When high-fluid intelligence systems start to show up, they will immediately take over the knowledge-dependent ones. Because they will be able to scale their knowledge just as well as legacy systems (knowledge gathering is the easy part), while their ability to recombine and