OpenAI Launches ChatGPT Health to Offer Personalised Responses Based on Apple Health and Medical Records

OpenAI introduced ChatGPT Health in January as a dedicated workspace for users, where they can ask health-related questions to the AI-powered chatbot. The ChatGPT Health functionality has been in early access since its unveiling, with limited users onboarded through a waitlist. Now, the Sam Altman-led AI giant has announced that ChatGPT Health is now generally available to all users in the US. The company says that users will have to connect their health apps, like Apple Health, and provide their medical records to help the AI agent generate personalised responses. OpenAI also highlighted that ChatGPT users have started asking more health-related questions every week.

ChatGPT Health Now Generally Available in the US

On Thursday, the US-based AI giant launched ChatGPT Health, which is now generally available to all logged-in free and paid users aged 18 and above in the US on web and iOS. However, the company also highlighted that “Health is not yet available in Codex”. OpenAI says that users do not have to connect other apps to ask health-related questions to ChatGPT.

However, users who want personalised, contextual answers to their health queries will have to “securely” connect their Apple Health app and “supported” medical records from different hospital systems in the US, including One Medical and Function Health. Users can access ChatGPT Health from the sidebar or the More menu. ChatGPT will guide users to connect their health records and metrics with the AI chatbot.

ChatGPT Health will be able to access a user's medications, lab results, recent visits, and sleep activity from the connected records, while also considering their health goals and the context they provide. As an example, the company said that ChatGPT Health will be able to consider dietary restrictions when a user is looking for a restaurant. The AI chatbot will also be able to consider recent injuries “when planning lower-impact weekend activities”. Users will also be able to manage their health-related information, view data and trends, browse synced health records, and return to past health conversations.

The Sam Altman-led AI giant also highlighted that its recently launched GPT-5.6 Sol AI model outperformed the GPT-5.5 Instant and GPT-4o models on the HealthBench Professional platform, which evaluates AI performance on health risks, based on accuracy, safety, communication, context awareness, completeness, and appropriate escalation. The GPT-5.6 Sol model scored 91, 86.7, 88, 93.8, and 83 points in the Accuracy, Communicates Clearly, Completeness, Follows Instructions, and Health Decision Helpfulness tests.

Further, the US-based tech firm claims that ChatGPT users ask more than 300 million health-related questions to the AI chatbot every week. OpenAI also claims that the AI agent is being used to understand lab results, prepare for appointments with doctors, interpret what a doctor is saying, and build a “healthier routine”.