The June 2026 update to GPT-5.5 Instant targeted how ChatGPT handles a request that develops over several messages. OpenAI's June 24 release notes describe better retention of the user's goal and constraints, and more effective adaptation when the user corrects or challenges an answer.
That gives concrete meaning to the brief announcement about an updated model: the intended improvement was keeping recommendations aligned with the evolving conversation. The notes describe expected behavior rather than publishing a measured improvement rate.
