Making ChatGPT better for clinicians

Making ChatGPT better for clinicians | OpenAI

April 22, 2026

Making ChatGPT better for clinicians

Built for clinical work, ChatGPT for Clinicians is now available for free to verified individual clinicians in the U.S.

Get started Learn more about HealthBench Professional

Loading…

Share

We’re introducing ChatGPT for Clinicians, a version of ChatGPT designed to support clinical tasks like documentation and medical research so clinicians can focus on delivering high-quality patient care. We’re making it free for any verified physician, NP, PA, or pharmacist, starting in the U.S.

The U.S. healthcare system today is under extraordinary strain. Clinicians are being asked to care for more patients while managing growing administrative demands and a rapidly expanding body of medical research. Many are already turning to AI tools like ChatGPT for support. According to a 2026 survey by the American Medical Association⁠, physician use of AI is now at an all-time high, with 72% of physicians reporting they now use AI in clinical practice, up from 48% last year. Today, millions of clinicians worldwide use ChatGPT to support their clinical care every week, for applications like care consult, writing and documentation, and medical research. Clinician usage of ChatGPT has more than doubled over the past year.

As demand for AI in clinical settings grows, so does the responsibility to continuously improve our model’s performance and safety on clinical use cases and offer solutions that can safely and effectively support healthcare workflows. Earlier this year, we introduced ChatGPT for Healthcare,⁠ which allows organizations to deploy ChatGPT to clinicians, administrators, and researchers with the compliance and controls they need at scale. Clinicians across leading U.S. health systems are now using it to move faster through administrative work like medical research and documentation, and get time back for patient care.

Enabling free access to ChatGPT for Clinicians is the next step, in support of our mission to ensure AGI benefits all of humanity. Learn more on our website⁠ or get started⁠.

ChatGPT for Clinicians builds on our foundation of continual model evaluation and improvement in health in partnership with clinicians. With its release, we are also introducing HealthBench Professional⁠, an open benchmark for real clinician chat tasks across three use cases: care consult, writing and documentation, and medical research, building on HealthBench’s⁠ broader evaluation of health conversations.

> “This version of ChatGPT is as close to an ideal clinical support partner as it gets. It’s like an on-demand consultant I can engage on everything from current guidelines to billing and coding, with the added benefit of broad access to pediatric and pediatric subspecialty literature.”

Designed for and with clinicians

Our team worked with hundreds of physician advisors to inform and improve capabilities for ChatGPT for Clinicians, and ensure they support key clinician use cases.

ChatGPT for Clinicians includes:

• Advanced AI models for complex clinical questions: Free access to our current frontier models for healthcare use cases—to help handle questions, research, and documentation more reliably.
• Skills for repeatable clinical workflows: Turn common workflows into reusable skills so ChatGPT can follow the same steps each time for tasks like referral letters, prior auth, and patient instructions.
• Trusted clinical search: Reason through cases faster and with greater confidence with real-time, cited answers based on evidence from millions of reputable, peer-reviewed medical sources.
• Deep research across medical journals: Delegate medical literature reviews to ChatGPT, set the sources you trust, steer the research if needed, and let it compile a comprehensive, well-cited report in minutes.
• CME from real clinical questions: As you research clinical questions in ChatGPT, eligible evidence review can automatically count toward continuing medical education credits—without separate courses or extra paperwork.
• Optional support for HIPAA compliance: Many clinical tasks don’t require PHI, but if needed, HIPAA support is available through a Business Associate Agreement (BAA) for eligible accounts.
• Account security and privacy: Conversations are not used to train models, plus protections like multi-factor authentication, help keep sensitive work secure.

Learn more about ChatGPT for Clinicians⁠ or get started⁠.

Continuing to evaluate and strengthen model health performance and safety

We are always improving the safety and accuracy of ChatGPT’s responses in health scenarios. OpenAI’s physician advisors continuously review model responses and provide feedback on quality, reasoning, trustworthiness, and safety. To date, they have reviewed more than 700,000 model responses that reflect how clinicians and patients may use ChatGPT in the real world; every few minutes a new model response is reviewed by a physician.

OpenAI’s models rank as the top-performing systems for real-world healthcare use on third party evaluations like Stanford’s MedHELM⁠ and MedMarks⁠. ChatGPT for Clinicians builds on models like GPT‑5.4, which outperform other models on OpenAI’s HealthBench⁠.

That rigor also shaped the development of ChatGPT for Clinicians. Before release, physician advisors tested 6,924 conversations in their daily work across clinical care, documentation, and research. Overall, physicians rated 99.6% of responses as safe and accurate. On a subset of 355 examples where for each, three independent physicians specified ground-truth citations, ChatGPT for Clinicians cited those sources more often than human physicians. Even so, ChatGPT for Clinicians is designed to support clinicians with information, not replace their judgment or expertise.

Today, we are also introducing HealthBench Professional⁠, an open benchmark for real clinician chat tasks across three use cases: care consult, writing and documentation, and medical research. Building on HealthBench’s broader evaluation of health conversations, it uses physician-authored conversations and rubrics, multi-stage physician adjudication, and careful data filtering to measure performance and safety in common clinician chats.

HealthBench Professional examples were chosen for their quality, representativeness, and difficulty to enable continued measurement of progress. About a third of examples involved physicians deliberately “red teaming,”…