GPT-6 Astra represents a step-function change in interactive reasoning

ARC Prize reported that OpenAI's GPT-6 Astra scored 66% on the ARC-AGI-3 benchmark using a standard harness, and nearly 100% with a continuous conversation harness and custom compaction at a cost of roughly $360 per game. The model outperformed human baselines in action efficiency and demonstrated on-the-fly symbolic world modeling, including developing its own shorthand DSL, marking a step-function change in interactive reasoning.

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: