How lfnovo/open-notebook works
Open Notebook is a self-hosted research workspace for collecting and organizing sources, then using configurable AI models to search, chat, transform content, and generate podcasts. The principal workflow is user → web interface/API → source processing and notebook storage → embedding/search or contextual chat → cited research results. The map also includes model configuration and podcast generation as major capabilities. Several subsystem links are documented in the README; where source excerpts do not expose runtime wiring, those links are marked unverified or omitted.
Main entry points: notebook.py Notebooks and notes · routers Feature routes · models.py Model manager · embedding_commands.py Embedding jobs
Components 20
Web experience
- Notebook UI
frontend/src/app/(dashboard) - API client
frontend/src/lib/api
API and jobs
- REST API
api/main.py - Feature routes
api/routers - Job submission
api/command_service.py
Research workflows
- Source processing
open_notebook/graphs/source.py - Source jobs
commands/source_commands.py - Embedding jobs
commands/embedding_commands.py - Search and ask
open_notebook/graphs/ask.py - Research chat
open_notebook/graphs/chat.py - Transformations
open_notebook/graphs/transformation.py
AI and media
- Model manager
open_notebook/ai/models.py - Provider registry
open_notebook/ai/provider_registry.py - Podcast generation
commands/podcast_commands.py - Podcast service
api/podcast_service.py
Domain and storage
- Notebooks and notes
open_notebook/domain/notebook.py - Repository layer
open_notebook/database/repository.py
Other components
- Researcher
- SurrealDB
- AI providers
Connections 22
- Researcher → Notebook UI: uses
README.md - Notebook UI → API client: requests data
- API client → REST API: calls API
- REST API → Feature routes: routes requests
api/main.py - Feature routes → Job submission: submits jobs
api/routers/embedding.py - Feature routes → Notebooks and notes: manages records
api/routers/embedding.py - Feature routes → Search and ask: invokes ask flow
api/routers/search.py - Feature routes → Podcast service: requests podcasts
- Source processing → Notebooks and notes: uses domain records
open_notebook/graphs/source.py - Source jobs → Source processing: runs processing
- Job submission → Embedding jobs: submits embedding jobs
api/routers/embedding.py - Embedding jobs → Notebooks and notes: loads sources and notes
commands/embedding_commands.py - Embedding jobs → Model manager: generates embeddings
commands/embedding_commands.py - Search and ask → Notebooks and notes: retrieves research context
- Research chat → Notebooks and notes: uses notebook context
- Transformations → Notebooks and notes: transforms content
- Podcast service → Podcast generation: submits generation
api/podcast_service.py - Podcast generation → Model manager: resolves models
commands/podcast_commands.py - Model manager → Provider registry: uses provider setup
- Model manager → AI providers: calls models
- Notebooks and notes → Repository layer: loads and saves
open_notebook/domain/notebook.py - Repository layer → SurrealDB: queries records
open_notebook/database/repository.py