OpenAI claims that ChatGPT’s new default model is much less hallucinatory

OpenAI’s latest virtualization model for ChatGPT may not configure things as much. She has hallucinations It was an ongoing problem As for AI models, but OpenAI says its new GPT-5.5 Instant model has “significant improvements in realism across the board.”

The company claims that based on “internal evaluations,” GPT-5.5 Instant generated “52.5% fewer hallucination claims” than its GPT-5.3 Instant model “in high-stakes claims covering areas such as medicine, law, and finance.” GPT-5.5 Instant also “reduced inaccurate claims by 37.3% in particularly difficult conversations that users flagged for factual errors.”

OpenAI also claims that GPT-5.5 Instant is “more capable of performing everyday tasks,” such as analyzing image uploads and knowing when to turn to the web for an answer. GPT-5.5 Instant has “tighter, more precise” responses and will avoid the use of “gratuitous emojis.”

With GPT-5.5 Instant, ChatGPT is now “more efficient” at pulling context from things like past chats and your Gmail to give you more personalized responses, too. (This is a feature provided by Google Invest heavily in Gemini Also.) For all ChatGPT forms, the new Memory Sources feature will allow the chatbot to show what context was used to inform custom responses, and you can delete or correct the information if you need to.

OpenAI will begin rolling out GPT-5.5 Instant on Tuesday “to all ChatGPT users,” though GPT-5.3 Instant will be an option for three months until it “retires.” (In the past, users They mourned the loss of old modelsThis gives people time to move in.)

The enhanced customization will roll out first to Plus and Pro users on the web and will soon be available for mobile apps. OpenAI has “plans” to offer it “soon” to Free, Go, Business, and Enterprise users. The memory resources feature in ChatGPT consumer plans is rolling out now on the web and “coming soon to mobile.”