OpenAI has officially launched ChatGPT for Clinicians, a specialized version of its AI assistant tailored to support the administrative and research demands of medical professionals. The platform, which has seen a massive surge in usage over the past year—with 72% of physicians now reporting AI usage in practice—is now available for free to verified physicians, nurse practitioners (NPs), physician assistants (PAs), and pharmacists in the U.S.

Designed for Clinical Workflows

The tool was developed in close partnership with hundreds of physician advisors to ensure it handles medical tasks with accuracy and professional rigor. Key features include:

• Advanced Medical Reasoning: Powered by models like GPT-5.4, which outperform other models on OpenAI’s clinical benchmarks.

• Repeatable Clinical Skills: Create reusable workflows for tasks like drafting referral letters, managing prior authorizations, and generating patient instructions.

• Evidence-Based Search: Provides cited answers based on millions of reputable, peer-reviewed medical sources, helping clinicians research cases with higher confidence.

• CME Credits: Clinicians can earn Continuing Medical Education (CME) credits by performing evidence-based research directly within the platform.

• Security & Compliance: Conversations are not used to train models. Optional HIPAA support is available via a Business Associate Agreement (BAA) for eligible institutional accounts.

Safety and Performance

OpenAI has placed a massive emphasis on health safety, with physician advisors reviewing over 700,000 model responses to ensure quality and reasoning.

• HealthBench Professional: OpenAI is also launching this new open benchmark to measure performance on "care consult," "writing/documentation," and "medical research" tasks.

• Evaluation Results: In internal evaluations, GPT-5.4 within the clinical workspace outperformed base GPT-5.4, all external models, and even human-written physician responses in certain specialized medical scenarios.

• Human-in-the-Loop: OpenAI emphasizes that ChatGPT for Clinicians is an information tool, not a replacement for clinical judgment or human expertise.

Global Expansion and Access

The free version is currently limited to the United States. However, OpenAI plans to expand access internationally in the coming months, beginning with a pilot program through the Better Evidence Network. The company has also released a Health Blueprint providing recommendations for the responsible integration of AI in U.S. healthcare.

This release is a landmark example of domain-specific AI. By tailoring the interface and reasoning capabilities to clinical workflows, OpenAI is demonstrating how "frontier models" can be constrained and tuned for high-stakes, professional environments where accuracy, attribution, and data security are the baseline for adoption. Verified clinicians can get started on the OpenAI website.