According to Voyant Beating monitoring, OpenAI has released GPT-5.5 Instant, a major upgrade to the default model of ChatGPT, now available to all users. GPT-5.5 Instant has seen comprehensive improvements in intelligence, fact-checking accuracy, and response quality. Key enhancements include a significant increase in fact-checking accuracy, with a 52.5% reduction in hallucinations in high-risk areas such as medicine, law, and finance, and a 37.3% reduction in overall inaccurate statements. Additionally, the model has been enhanced in areas such as STEM queries, image analysis, tool integration (e.g., web search), and visual reasoning. The response style has also been optimized—answers are more concise (with a reduction of around 30% in word count), reducing redundant formatting, excessive emojis, and unnecessary follow-up questions, maintaining a warm personality while avoiding clutter.

The new model excels in personalization, intelligently leveraging chat history, uploaded files, connected Gmail, and other contexts to provide highly tailored suggestions and plans, while ensuring users retain full control. GPT-5.5 Instant exhibits a more natural and practical tone in everyday conversations, such as offering concise workplace advice or personalized recommendations based on user preferences (e.g., tea shop choices). OpenAI emphasizes that since Instant serves as a "day-to-day driver" assistant for most people, these subtle yet comprehensive improvements significantly enhance the user experience, making ChatGPT a more reliable and efficient AI partner.