Open Health
AI Health Assistant / Personal Health RecordTypeScript / Next.js / ReactAGPL-3.0

Open Health Review

A Next.js self-hosted AI health assistant with PostgreSQL/Prisma for structured health data and multi-provider AI chat.

Deployability
3/5
Value
4/5
Privacy
4/5

Each review covers deployability, value versus commercial alternatives, and privacy model. Tools that can run locally were started and exercised; mobile or backend-dependent tools were assessed from published builds, source code, and deploy guides. Ratings reflect what we were able to verify.

Open Health — Open Source AI Health Assistant

A Next.js 15 + React 19 self-hosted health data platform with PostgreSQL/Prisma, multi-provider AI chat, and support for lab reports, checkups, physical info, family history, and symptoms. Last updated: 2026-06-21.

One-sentence verdict: A polished, actively maintained open-source health assistant with real database-backed data management and multi-model AI chat — the strongest self-hosted alternative to commercial health Q&A apps in this batch.


What the System Is

Open Health is a web-based AI health assistant. The repository and deploy guide show:

  • Next.js 15 + React 19 + TypeScript frontend with Tailwind CSS and shadcn/ui.
  • PostgreSQL + Prisma for structured health data storage.
  • next-auth v5 credentials-based authentication.
  • Multi-provider AI: OpenAI, Anthropic, Google, Ollama, and other compatible providers.
  • Health data inputs: blood test results, health checkup data, physical information, family history, and symptoms.
  • Two modes: Clinic (quick Q&A) and full platform (comprehensive management).

The deploy guide states the service was started on http://localhost:3000 and the screenshot was captured from the running instance.

Key data
Category AI Health Assistant / Personal Health Record
Language TypeScript / Next.js / React
License AGPL-3.0
Self-hosted Yes
AI OpenAI, Anthropic, Google, Ollama, others
Database PostgreSQL + Prisma
Deployment Node.js dev server

How to Install and Deploy

Prerequisites: local PostgreSQL.

cd /data2/docker/going-global/repos/open-health
npm install
npx prisma generate
npx prisma migrate dev --name init
npm run dev

Required environment variables in .env:

DATABASE_URL=postgres://postgres:mysecretpassword@127.0.0.1:5432/open-health
AUTH_SECRET=<random-string>
AUTH_TRUST_HOST=true
AUTH_URL=http://localhost:3000

Open http://localhost:3000, sign up with any email format (local mail is not configured), and start using the platform.


How to Test

The documented test flow is:

  1. Verify http://localhost:3000/login loads.
  2. Sign up with an email and password; the app auto-logs you in.
  3. Explore the chat interface in Clinic or Q&A mode.
  4. Add a health data source: blood test, checkup, physical info, family history, or symptom.
  5. Verify structured data appears and can be used as context for AI chat.
  6. Test provider switching if API keys are configured.

Privacy & Compliance

Open Health is designed for personal health data, not covered-entity PHI. It is not advertised as HIPAA compliant. Self-hosting keeps data on your own PostgreSQL instance, which is a strong privacy foundation, but compliance depends on the host environment, access controls, backups, and AI provider configuration. Do not store others' PHI without a proper compliance review.


Open Health vs Commercial Health Assistants

Dimension Open Health Commercial Health Assistant (e.g., Babylon, K Health, Ada Health)
Cost Free / self-hosted Freemium or subscription
Data location Your own PostgreSQL database Vendor cloud
Open source Yes (AGPL-3.0) No
AI provider choice Multiple providers + local Ollama Vendor-controlled
Health data types Labs, checkups, physical info, family history, symptoms Varies
Setup effort Medium: PostgreSQL + Prisma + auth Low: sign up
Clinical validation Not claimed Often regulated/marketed as medical devices

Who Should Use It

  • Users who want a self-hosted, database-backed health assistant.
  • People comfortable with PostgreSQL and Next.js deployment.
  • Privacy advocates who prefer local AI providers like Ollama.

Who Shouldn't Use It

  • Users who need a turnkey mobile app or provider-integrated records.
  • People expecting clinically validated diagnosis or triage.
  • Anyone storing PHI for patients without a compliance review.

FAQ

Is Open Health free?

Yes. It is open source under AGPL-3.0 and free to self-host.

Can I run Open Health without sending data to OpenAI?

Yes. The settings support local providers such as Ollama, so AI inference can stay on your machine.

Does it connect to my doctor or lab?

No. Data must be entered manually or imported from files. There is no direct EHR or lab integration in the open-source version.


Verdict

Open Health is a polished, actively maintained open-source health assistant. It combines structured health data management with multi-provider AI chat in a modern Next.js stack. The PostgreSQL backend and credentials auth make it more serious than most prototypes in this batch. The main limitations are manual data entry and the need for self-hosted infrastructure.

Ratings: Deployability 3/5 · Value vs Commercial 4/5 · Privacy Compliance 4/5