tech
Improving health intelligence in ChatGPT
GPT‑5.5 Instant brings frontier health intelligence to more people, shaped by advances in our models and physician-led evaluation.
TL;DR
- GPT‑5.5 Instant brings advanced health intelligence to all free ChatGPT users.
- The model shows improvements in recognizing urgent care needs, asking for context, explaining uncertainty, and simplifying complex health information.
- Performance on challenging health evaluations is now comparable to OpenAI's frontier Thinking models.
- Physician-led evaluations, including HealthBench and HealthBench Professional, are key to measuring and improving health responses.
- GPT‑5.5 Instant responses were rated higher than physician-written responses and older model responses in a comparative study.
- Production traffic data shows a 71% reduction in factuality issues in health responses over the last two months.
- OpenAI collaborates with over 260 physicians globally to refine health responses across various specialties and languages.
- Physician feedback informs evaluation criteria, leading to more accurate, safe, clear, complete, and useful health guidance.
- These health improvements also support other OpenAI healthcare tools like ChatGPT for Clinicians.