Get 125% / $2,500 on 1st deposit!
New players only. Exclusive Welcome Bonus of up to $2,500
OpenAI released GPT-5.5 Instant on Tuesday as the new default ChatGPT model, replacing GPT-5.3 Instant while keeping the same low-latency focus.
Good to Know
OpenAI is changing the default ChatGPT model again, but the company wants users to see the upgrade as practical rather than disruptive. GPT-5.5 Instant keeps the fast response style of GPT-5.3 Instant while improving reasoning, coding and knowledge work.
The benchmark gains are clear. GPT-5.5 Instant reached 81.2 on the AIME 2025 math test, compared with 65.4 for the older Instant model. It also scored 76 on MMMU-Pro, above the 69.2 result from GPT-5.3 Instant.
Context handling may affect daily users more than benchmark scores. GPT-5.5 Instant can use search tools to refer back to past chats, files and Gmail, giving more personal answers when access is available. Plus and Pro users on web get that feature first, while mobile access will follow. OpenAI also plans wider access for Free, Go, Business and enterprise users in the coming weeks.
OpenAI will now show memory sources across all ChatGPT models. Users can check where an answer came from, delete outdated memory sources or correct wrong ones. Shared chats will not reveal those memory sources to other people.
Developers will get GPT-5.5 through the API as chat-latest. Paid users can still choose GPT-5.3 for three months before that option goes away.
Model changes have caused user frustration before. OpenAI faced backlash after retiring GPT-4o, which many users liked for its warmer style. Some petition supporters described it as their “best friend” or “a mirror.” OpenAI still deprecated GPT-4o in February 2026.