OpenAI introduced a new foundation model, GPT-5.5 Instant, set to replace GPT-5.3 Instant as the default model in ChatGPT.

The company on Tuesday said the upgrade reduces hallucinations in high-stakes areas such as law, medicine, and finance, while preserving the fast response times of the previous version.

GPT-5.5 was originally released last month, with OpenAI highlighting improvements in coding performance and general knowledge tasks, positioning it as a stronger model for professional and productivity use cases.

The new model also scored 81.2 on the AIME 2025 math benchmark, up from 65.4 for the previous version.

It likewise showed stronger multimodal reasoning on the MMMU-Pro test, scoring 76 compared to 69.2.

A key focus of the release is improved context management. GPT-5.5 Instant can use its search tool to draw on past conversations, files, and Gmail content to generate more personalized responses.

The feature will first be available to Plus and Pro users on the web, with mobile support coming later.

OpenAI also said it plans to extend access to Free, Go, Business, and enterprise users in the coming weeks.

With this update, ChatGPT will display memory sources across all models, allowing users to see where responses are drawn from. Users will also be able to remove outdated sources or correct them if they are inaccurate. Importantly, OpenAI noted that shared chats will not include visibility into these memory sources.

For developers, GPT-5.5 will be accessible via the API under the name “chat-latest,” while GPT-5.3 will remain available as an option for paid users for a limited three-month period.