Many of you will ask, "if it saturates ARC 3, is it AGI?" We're not making this claim. All we know about the system so far are its benchmark scores. When we launched ARC 3, and in every presentation we made about it, we were very insistent on one thing: solving it is not proof of AGI. It's not intended as a finish line. ARC 3 is testing the right qualitative properties you'd expect of an AGI system -- exploration under uncertainty, adaptation without instructions, causal world modeling from limited data, etc. -- but in small quantities. ARC 3 games are orders of magnitude shorter timescales than real world tasks, and represent orders of magnitude less data, less modeling complexity, less on-the-fly learning. (Slide below is from a March 2026 presentation)

GPT-6 Astra represents a step-function change in model capability for interactive reasoning problems. It scores 66% on ARC-AGI-3 using our standard harness, and nearly 100% with a continuous conversation harness and custom compaction, at a cost of roughly $360 per game. In fact, the continuous harness version significantly outperforms our human baseline in action efficiency across almost all levels. When we examined the reasoning chains to understand how the model operates, we found it performing highly efficient, on-the-fly symbolic world modeling for each game and level. It goes as far as developing its own shorthand DSL to represent in-game situations -- essentially a game-specific algebraic notation. Overall, Astra exhibits symbolic modeling behaviors we had previously only seen with sophisticated harnesses -- so harness capabilities are increasingly shifting into the model itself. We see Astra as a major breakthrough in model intelligence. Read our post on Astra and what these results mean: arcprize.org/blog/astra