Last updated: November 16, 2026
On November 16, 2026, OpenAI updated ChatGPT for Clinicians and ChatGPT for Healthcare. Clinical search now finds, evaluates and cites medical evidence better, computer-use agents run in ChatGPT Work and Codex, and more than 20 healthcare plugins arrived. ChatGPT for Clinicians stays free for verified U.S. clinicians.
Contents
- What changed on November 16
- Clinicians vs. Healthcare: who gets what
- How to sign up and connect plugins
- Improved trusted clinical search
- Computer-use agents and the models behind them
- New healthcare plugins
- Limitations and privacy rules
- Frequently asked questions
What changed on November 16
OpenAI announced three changes in its post More ways to move care forward with ChatGPT for Healthcare, and the ChatGPT release notes list “Clinical search and plugins for ChatGPT for Clinicians” under the same date:
- Improved trusted clinical search in both ChatGPT for Healthcare and ChatGPT for Clinicians, covering how ChatGPT finds, evaluates and cites medical evidence.
- Computer-use agents that read screens, click, type and move between applications such as EHRs and payer portals, available in ChatGPT Work and Codex.
- More than 20 new healthcare plugins for ChatGPT for Healthcare, with select plugins also available to individual clinicians.
For OpenAI’s related agent features in Enterprise, Edu and Healthcare workspaces, see our ChatGPT Dots guide.
Clinicians vs. Healthcare: who gets what
The two products share the new capabilities but differ in eligibility and management. This table draws on the announcement and the ChatGPT for Clinicians help article.
| Feature | ChatGPT for Clinicians | ChatGPT for Healthcare |
|---|---|---|
| Who it is for | Individual verified U.S. clinicians (MD/DO, NP, PA, pharmacists) | Healthcare organizations: clinical, research and administrative teams |
| Price | Free | Through OpenAI sales or an AWS Marketplace private offer |
| Improved clinical search | Available now | Available now, subject to workspace settings |
| Computer-use agents in Work and Codex | Available now | Available now; administrators enable it |
| New healthcare plugins | Select plugins | All 20+ new plugins, enabled by an administrator |
| BAA | Individual BAA under Settings > Agreements | BAA covering multiple users |
| Epic plugin | Not available | Available with organization setup |
OpenAI says content shared with either product is not used for model training.
How to sign up and connect plugins
You need a ChatGPT account, a valid NPI and a license that the third-party verification provider can confirm. The help article lists these steps:
- Go to the ChatGPT for Clinicians sign-up page and sign in with your ChatGPT account, or create one.
- Complete clinician verification with your NPI through the third-party provider.
- Confirm that you are a licensed clinician and accept the services agreement.
- Select Get started with ChatGPT to create your Clinicians workspace. Your existing workspace stays available in the account switcher.
- If you will enter protected health information, sign the BAA under Settings > Agreements first.
- Open the Plugin directory, select Healthcare Public Data, choose Install plugin, then select Connect for each app you want.
- Ask a question, or type
@in ChatGPT (or$in Codex) to query a specific connected app.
According to the Healthcare Public Data help article, installing the plugin does not connect its apps automatically. Each app keeps its own permissions, and there is no separate charge for the plugin.
Improved trusted clinical search
Trusted clinical search draws on peer-reviewed studies, public health guidance and clinical guidelines. Answers include citations with titles, journals, authors and publication dates, and ChatGPT decides which sources to use based on the prompt. Some journal articles still require a subscription or institutional access.
OpenAI reports that in a blind evaluation of real-world clinical questions, including multi-turn conversations, physicians rated the updated experience higher on all eight dimensions measured across more than 1,300 reviews. The dimensions included factual correctness, safety, actionability and source authoritativeness. This is OpenAI’s own evaluation, not an independent study.
Computer-use agents and the models behind them
The companion post, Advancing computer-use agents for healthcare teams, explains the agent side. Agents complete multi-step work in software that lacks a usable API, such as preparing a prior authorization in an insurer’s portal from records in the EHR. A person reviews the drafts before anything is submitted.
OpenAI published these results for two models:
- GPT-6 Astra on EHR inbox tasks: 99% of 325 reviewed tasks rated clinically safe, and physicians chose “very helpful” or “extremely helpful” in 93% of 646 assessments.
- HealthAdminBench, a Stanford benchmark of administrative workflows: GPT-6 Astra earned 94.3% of rubric points and GPT-6 Luna earned 91.8%, with Luna costing 98% less at 8.5 cents per task. Both ran at maximum reasoning effort.
Individual clinicians can try computer use in ChatGPT Work and Codex from the Clinicians workspace. If Work or Codex shows an out-of-credits message, the help center says to switch to the Clinicians workspace first; credits bought in another workspace do not carry over. Developers can build similar agents with the OpenAI API; our OpenAI Agents API guide and Codex CLI 0.169 guide cover the developer tools, and our Gemini Computer Use guide describes Google’s competing computer-use tool.
New healthcare plugins
The 21 plugins named in the announcement are Aiva, Amion, Assort Health, Availity Extend, Elation Health, Ferry Health, IMO Health, Infinitus Relay, K Health, Luma Health, Metriport, Notable, Oracle HDI, PenguinAI, Photon, Redox, Stedi, Subsalt, TriZetto, Turquoise and UpToDate Expert AI. They cover patient records, hospital protocols, prescriptions, outreach, insurance eligibility and cost analysis.
For individual clinicians, the help article describes two plugin options. Healthcare Public Data bundles nine read-only apps: PubMed, ClinicalTrials.gov, DailyMed, RxNorm, openFDA, CMS Coverage, CMS Open Data, Medicare Care Compare and the NPI Registry. The Photon plugin lets verified clinicians draft, review and send prescriptions through a connected Photon account, with each prescription authorized individually. Controlled substances are not supported.
Limitations and privacy rules
- U.S. only. ChatGPT for Clinicians is limited to verified clinicians in the United States.
- No PHI in public-data searches. Do not put patient names, record numbers or other identifiers into Healthcare Public Data queries; a BAA does not by itself authorize an external service to receive PHI.
- No chart access for individuals. Public-data apps do not retrieve patient charts, and Epic is not available in Clinicians workspaces.
- Plugin access varies. Availability depends on plan, workspace, region and each plugin’s requirements, and third-party plugins have their own terms.
- Vendor-reported results. The safety and benchmark figures come from OpenAI’s own evaluations.
- Clinical judgment still applies. OpenAI states the product supports, not replaces, professional judgment.
OpenAI has used similar approval-first designs elsewhere, such as purchase approvals in ChatGPT Wallet.
Frequently asked questions
What is new in ChatGPT for Clinicians?
On November 16, 2026, OpenAI improved trusted clinical search, made computer-use agents in ChatGPT Work and Codex available to individual clinicians, and added access to select healthcare plugins, alongside more than 20 new plugins for ChatGPT for Healthcare.
Is ChatGPT for Clinicians free?
Yes. OpenAI says ChatGPT for Clinicians is free for verified clinicians in the United States. Eligible clinicians currently include physicians (MD/DO), nurse practitioners, physician assistants and pharmacists.
Does ChatGPT for Clinicians offer a BAA for HIPAA?
Yes. Eligible individual clinicians can review and sign a Business Associate Agreement under Settings > Agreements. OpenAI says not to share protected health information unless a BAA is in place.
Can ChatGPT for Clinicians access Epic patient records?
No. OpenAI's help center says the separate Epic plugin is not available with ChatGPT for Clinicians. Epic access requires an organizational ChatGPT for Healthcare setup and an individual Epic sign-in.
What is the difference between ChatGPT for Clinicians and ChatGPT for Healthcare?
ChatGPT for Clinicians is a free, self-serve product for individual U.S. clinicians. ChatGPT for Healthcare is for organizations that need centralized administration, organization-wide BAAs, company data connections and enterprise controls.
Which models power the healthcare computer-use agents?
OpenAI's evaluation names GPT-6 Astra and GPT-6 Luna. On HealthAdminBench, Astra earned 94.3% of rubric points and Luna 91.8% at 8.5 cents per task, both at maximum reasoning effort.
Sources
- More ways to move care forward with ChatGPT for Healthcare, OpenAI (November 16, 2026)
- Advancing computer-use agents for healthcare teams, OpenAI (November 16, 2026)
- ChatGPT for Clinicians, OpenAI Help Center
- Using Healthcare Public Data in ChatGPT and Codex, OpenAI Help Center
- ChatGPT release notes, OpenAI Help Center
Featured image: OpenAI logo, public domain, via Wikimedia Commons. OpenAI and ChatGPT are trademarks of OpenAI. FreeAnswers is not affiliated with OpenAI. This guide is not medical or legal advice.