OpenAI Unveils GPT-5.5: Less Hallucination, More Accuracy
OpenAI's new GPT-5.5 Instant model promises to reduce hallucinations significantly. Discover how this upgrade enhances ChatGPT's accuracy and personalization features.
Major Improvements in GPT-5.5 Instant
OpenAI has announced its latest default model for ChatGPT, the GPT-5.5 Instant, which aims to tackle the persistent issue of AI hallucinations. According to the company, this new model has shown a remarkable 52.5% reduction in hallucinated claims compared to its predecessor, GPT-5.3, particularly in high-stakes areas like medicine, law, and finance.
In addition to improved factuality, GPT-5.5 Instant offers enhanced capabilities for everyday tasks. Users can expect:
- •Tighter, more concise responses
- •Reduced use of unnecessary emojis
- •Better context awareness from previous chats and emails
OpenAI is rolling out this model to all ChatGPT users, with Plus and Pro users receiving the enhanced personalization features first. The transition period allows users to adapt to the new model while still having access to GPT-5.3 for three months.