OpenAI Releases GPT-5.5 Instant for ChatGPT With Improved Accuracy and Memory Controls

OpenAI has released GPT-5.5 Instant for ChatGPT, replacing GPT-5.3 Instant as the default model and adding new memory-source controls.

In a product post dated May 5, OpenAI said GPT-5.5 Instant is rolling out starting that day and is meant to improve factuality and make responses more concise. The company also said the update is designed to feel more personalized.

A key addition is memory-source visibility. OpenAI says users can now see what context is being used for personalized responses and can delete or correct that context.

OpenAI's help center says the model is the default for logged-in users, while reporting from TechCrunch, The Verge, Axios and MacRumors all tracked the same rollout.

The practical significance is straightforward: ChatGPT's default experience is changing again, with more emphasis on accuracy, tighter answers and user control over the context feeding the model.

The immediate question is how quickly the update reaches all account tiers and devices, and whether OpenAI expands or adjusts the rollout as users start testing the new model.