OpenAI's GPT-5.5 Instant brings substantial improvements to ChatGPT's health and wellness capabilities. Over 230 million people use ChatGPT weekly for health questions, from understanding lab results to navigating insurance. The new model better recognizes urgent care needs, asks for relevant context, and explains uncertainty more clearly than previous versions.

Performance now matches OpenAI's frontier Thinking models on challenging health evaluations. This improvement comes from physician-led development - 260 physicians across 60 countries and 26 specialties have reviewed over 700,000 model responses. Health Bench and Health Bench Professional evaluations assess accuracy, safety, communication, and appropriate escalation using realistic health conversations and physician-written rubrics.

In production traffic, the rate of health responses with flagged factuality issues dropped 71% in two months. Physicians rated GPT-5.5 Instant responses higher than both older models and physician-written responses across accuracy, communication, and completeness criteria. The model showed fewer failures in identifying red flags, tailoring to local healthcare contexts, and seeking necessary user context.

These upgrades benefit all free ChatGPT users, supporting OpenAI's broader healthcare tools including ChatGPT for Clinicians. The focus on physician expertise and rigorous evaluation suggests AI health assistance is reaching a new threshold of practical utility.