Try the hosted version here 🙌
Then fork and get building...
Enterprise-grade biomedical deep research - Pick a life sciences workflow, fill in a few variables, and get back a cited research report with deliverables. Powered by Valyu DeepResearch over PubMed, ClinicalTrials.gov, FDA drug labels, patents, and the open web.
Biomedical evidence is fragmented across dozens of databases, and stitching it together by hand is where the days go. Bio runs the whole search-read-synthesise loop for you:
- Prebuilt life sciences workflows - Competitive landscapes, clinical readouts, regulatory precedent, and business development scans, each with typed inputs instead of a blank prompt box
- Comprehensive biomedical data - PubMed articles, ClinicalTrials.gov records, FDA drug labels, patents, and more through Valyu's unified API
- Cited reports - Every claim carries an inline citation that resolves to the underlying source
- Deliverables - Structured artifacts alongside the narrative, plus PDF export
- Example reports - Real finished reports per domain, so you can see the output before spending a credit
- Self-hostable - Local SQLite, no authentication, no rate limits
- Workflow browser - Browse the life sciences workflow catalog by lens: Pipeline & Assets, Clinical, Regulatory, and Business Development
- Three research modes -
fast,standard, andheavy, trading depth against turnaround time - Live activity feed - Watch the searches, reads, and reasoning steps as the report is built
- Report history - Every report is stored and re-openable, with cancel and sync controls while a run is in flight
- Inline citations - Hover any marker for the source title, date, and link
- Images and deliverables - Charts and structured outputs rendered next to the report body
- PDF export - Server-rendered PDF of any completed report
- Share links - Send a report to a colleague or post it
Each domain ships with a seeded example report so the app is useful before you have run anything yourself. They render through the same report view as live output.
Self-hosted mode is the recommended way to run Bio. It provides a complete local environment with:
- No authentication required - Auto-login as dev user
- Local SQLite database - No external database setup needed
- No rate limits - Credits are handled by your own Valyu API key
- Node.js 18+
- pnpm (
npm install -g pnpm) - Valyu API key (get one at platform.valyu.ai)
-
Clone the repository
git clone https://github.com/yorkeccak/bio.git cd bio -
Install dependencies
pnpm install
-
Set up environment variables
Create a
.env.localfile in the root directory:# Enable Self-Hosted Mode NEXT_PUBLIC_APP_MODE=self-hosted # Valyu API Configuration (Required) VALYU_API_KEY=your-valyu-api-key # OpenAI Configuration (Optional - used to suggest report deliverables) OPENAI_API_KEY=your-openai-api-key # Local model servers (Optional - powers the local model status indicator) OLLAMA_BASE_URL=http://localhost:11434 # Default Ollama URL LMSTUDIO_BASE_URL=http://localhost:1234 # Default LM Studio URL
-
Run the development server
pnpm dev
-
Open your browser
Navigate to http://localhost:3000
You'll be automatically logged in as
dev@localhostwith full access to all features.
Self-hosted mode provides a complete local environment without any external dependencies beyond the Valyu API. It's perfect for:
- Local Development - No Supabase setup required
- Testing Features - No auth wall between you and the app
- Privacy - Reports and user data stay on your machine
- Quick Prototyping - No authentication or rate limits
When NEXT_PUBLIC_APP_MODE=self-hosted:
-
Local SQLite Database (
/.local-data/dev.db)- Automatically created on first run
- Stores the local user record
- Easy to inspect with
sqlite3 .local-data/dev.db
-
Mock Authentication
- Auto-login as dev user (
dev@localhost) - No sign-up/sign-in required
- Auto-login as dev user (
-
No Rate Limits
- Usage is governed by the credits on your own Valyu API key
Bio detects Ollama and LM Studio running on your machine and surfaces their status and available models in the top-right indicator.
Ollama:
# Install Ollama
brew install ollama # macOS
# OR
curl -fsSL https://ollama.com/install.sh | sh # Linux
# Start Ollama service
ollama serve
# Download a model
ollama pull qwen2.5:7bLM Studio:
- Download LM Studio from lmstudio.ai
- Download a model - Search for
qwen/qwen3-14borgoogle/gemma-3-12b - Start the server - Click the LM Studio menu bar icon -> "Start Server on Port 1234..."
- Configure the context window - Set to at least 8192 tokens (16384+ recommended)
View Database:
sqlite3 .local-data/dev.db
# Then run SQL queries
SELECT * FROM users;Reset Database:
rm -rf .local-data/
# Database recreated on next app startValyu mode adds hosted authentication and billing: users sign in with their Valyu account through OAuth, and research runs are charged against their Valyu credits rather than a shared API key.
- Valyu OAuth credentials (contact [email protected])
- A Supabase project for your app's own user data
# Enable Valyu Mode
NEXT_PUBLIC_APP_MODE=valyu
# Valyu OAuth Credentials (contact [email protected])
NEXT_PUBLIC_VALYU_SUPABASE_URL=https://your-valyu-supabase-url
NEXT_PUBLIC_VALYU_CLIENT_ID=your-client-id
VALYU_CLIENT_SECRET=your-client-secret
VALYU_APP_URL=https://platform.valyu.ai
# Your App's Supabase (for user data)
NEXT_PUBLIC_SUPABASE_URL=https://your-project.supabase.co
NEXT_PUBLIC_SUPABASE_ANON_KEY=your-anon-key
SUPABASE_SERVICE_ROLE_KEY=your-service-role-key
# Optional fallback key for anonymous traffic
VALYU_API_KEY=your-valyu-api-keyBio ships with a seeded example report for each life sciences lens:
- Pipeline & Assets - Competitive landscape across a target class
- Clinical - Trial landscape and readout analysis
- Regulatory - Approval pathway and precedent scan
- Business Development - Partnering and asset opportunity scan
Open one from the reports page to see the report view, citations, and deliverables without running a workflow.
- Frontend: Next.js 15 with App Router, Tailwind CSS v4, shadcn/ui
- Research: Valyu DeepResearch workflows over biomedical and web sources
- Auth: Valyu OAuth in valyu mode, local dev user in self-hosted mode
- Storage: Supabase in valyu mode, SQLite via Drizzle in self-hosted mode
- PDF: Puppeteer with
@sparticuz/chromium - Analytics: PostHog and Vercel Analytics
The quickest way to get Bio running in production:
- Fork this repository to your GitHub account
- Create a new project on vercel.com and import your fork
- Add environment variables in Vercel project settings (Settings > Environment Variables):
NEXT_PUBLIC_APP_MODE=self-hostedVALYU_API_KEY= your Valyu API keyOPENAI_API_KEY= your OpenAI API key (optional, for deliverable suggestions)
- Deploy - Vercel handles the rest
- Secure API key management
- No storage of sensitive medical data
- HTTPS encryption for all API calls
This project is licensed under the MIT License - see the LICENSE file for details.
Contributions are welcome! Here's how to get started:
- Fork the repository
- Create a feature branch (
git checkout -b feature/your-feature) - Make your changes
- Run
pnpm devand test locally - Commit your changes and push to your fork
- Open a Pull Request against
main
For bugs or feature requests, open an issue or start a discussion.
Made with love for biomedical researchers