OpenAI makes GPT-5.5 Instant ChatGPT’s new default model
OpenAI is swapping in GPT-5.5 Instant for ChatGPT’s default, promising 52.5% fewer hallucinations and faster, cleaner answers for millions.
OpenAI is putting GPT-5.5 Instant in ChatGPT’s default slot for everyone, betting that a quieter upgrade in everyday chats will matter more than flashy benchmark gains. The company says users should notice smarter and more accurate answers, clearer and more concise replies, a more natural conversational tone, and better use of the context already shared when personalization helps.
The biggest change for consumers is reliability. OpenAI said GPT-5.5 Instant produced 52.5% fewer hallucinated claims than GPT-5.3 Instant on high-stakes prompts and 37.3% fewer inaccurate claims on especially difficult conversations that users had flagged for factual errors. The model is also designed to improve photo and image uploads, STEM questions, and the system’s judgment about when web search can produce a better answer. OpenAI said the model now feels more dependable overall, with the largest gains in medicine, law and finance, areas where even small errors can carry outsized consequences.
Speed is part of the pitch too. OpenAI said GPT-5.5 matches GPT-5.4 per-token latency in real-world serving while being more capable, which means users should not have to trade responsiveness for better answers. The company framed Instant as the daily driver for hundreds of millions of people, a reminder that modest improvements can ripple across a huge share of ChatGPT traffic.
The broader GPT-5.5 launch, announced on April 23 and brought to the API the next day, shows where OpenAI is trying to push the product. The model is aimed at writing and debugging code, researching online, analyzing data, and creating documents and spreadsheets. In ChatGPT release notes, OpenAI said GPT-5.5 is built to understand complex goals, use tools, check its work and carry more tasks through to completion. GPT-5.5 Thinking is described as the most capable reasoning model in ChatGPT, while GPT-5.5 Pro is reserved for the hardest questions and highest-accuracy work.
OpenAI said GPT-5.5 is rolling out to Plus, Pro, Business and Enterprise users in ChatGPT and Codex, with GPT-5.5 Pro available to Pro, Business and Enterprise customers. On the API side, the model is priced at $5.00 per million input tokens and $30.00 per million output tokens, with higher charges for prompts above 272K input tokens. The company said it tested the model with internal and external red-teamers and nearly 200 trusted early-access partners before release.
The default-model switch also reflects the market pressure OpenAI faces. It has repeatedly used defaults to reshape the ChatGPT experience, moving to GPT-5 in August 2025 and then GPT-5.1 Instant in November 2025. The new handoff to GPT-5.5 Instant suggests the race now centers on the model most people never choose manually, but notice immediately when it gets answers wrong, or finally gets them right.
Know something we missed? Have a correction or additional information?
Submit a Tip
