ReadPilot

ReadPilot is a self-hosted PDF reader that integrates LLM capabilities directly into your reading workflow.

Unlike generic chat-with-PDF tools, ReadPilot automatically syncs the chat context with your current viewport. As you scroll, the AI "reads" along with you, allowing for precise questions without manual context management.

Quick Start (Docker)

Requires Docker Compose.

1. Configure Environment

   Copy the example config and set your API key (or use UI settings):

   cp .env.example .env
   # Optional: Set OPENAI_API_KEY, ANTHROPIC_API_KEY, or GEMINI_API_KEY

2. Run

   docker compose up --build

3. Open Browser

   • App: http://localhost:5173
   • Settings: Click the gear icon in the app to configure models or base URLs.

Data is persisted in the readpilot_data Docker volume.

Local Development

Backend (Python)

Requires Python 3.13+ and uv.

cd backend
uv sync
uv run uvicorn main:app --reload --port 8000

Frontend (React)

Requires Node.js 20+.

cd frontend
npm install
npm run dev

Configuration

Variable	Description	Default
OPENAI_API_KEY	OpenAI API Key	-
ANTHROPIC_API_KEY	Anthropic API Key	-
GEMINI_API_KEY	Google Gemini API Key	-
OPENAI_BASE_URL	Custom endpoint (e.g. for LocalAI/Ollama)	https://api.openai.com/v1
READPILOT_DATA_DIR	PDF & Metadata storage path	./storage

Note: ReadPilot requires models with native PDF input support (e.g., models from OpenAI, Anthropic, or Google Gemini).

License

MIT