LibreAI / RAG et connaissances / Open Notebook
// rag · lfnovo/open-notebook

Open Notebook

Open Notebook est un projet de RAG et connaissances open source, créé par lfnovo (lfnovo/open-notebook sur GitHub). Il compte 41 792 étoiles sur GitHub et 4 805 forks, il est publié sous la licence MIT et a été mis à jour pour la dernière fois sur GitHub le 8 novembre 2026.

Self-hosted NotebookLM alternative — private research notebooks, 20+ AI providers and multi-speaker podcasts, MIT

★ 41,8 k étoiles⑂ 4,8 k forksMITmis à jour nov. 2026
▦ visuel — le README de ce projet ne contient aucune capture intégrable,
voici donc un emplacement réservé. Le guide ci-dessous est complet malgré tout.
// détails du projet

Détails du projet

Étoiles GitHub
41 792
Forks
4 805
Licence
MIT
Dernière mise à jour sur GitHub
8 novembre 2026
Type de fiche
Guide d'installation complet
Site officiel
Official site
// fonctionnement

Comment ça marche

  1. Download the compose file

    Fetch the official docker-compose.yml into a folder.

  2. Set your encryption key

    Change OPEN_NOTEBOOK_ENCRYPTION_KEY to a long random secret — it encrypts the API keys you store, so keep it.

  3. Start it and connect a provider

    Bring the two containers up, open the UI, and add a provider (cloud API key, or a local Ollama base URL) under Models.

// fonctionnalités

Pourquoi on l'adore

20+ AI providers, or fully local

OpenAI, Anthropic, Google, Mistral, Ollama, LM Studio and any OpenAI-compatible endpoint — swap providers per model type.

Multi-modal sources

PDFs, videos, audio, web pages, Office docs, EPUBs and Reddit, all searchable in one notebook.

Citations on every answer

Chat answers cite the sources, notes and insights they used.

Multi-speaker podcasts

Professional podcast generation with up to 4 speakers and custom profiles — NotebookLM itself stops at 2.

Full REST API

Programmatic access to notebooks, sources and chat for custom integrations.

Privacy-first by design

Self-hosted; with local models nothing leaves your machine.

// lancez-le

Lancez-le vous-même

  1. Step 1 — Download the compose file

    Fetch the official docker-compose.yml from the repo.

    curl -o docker-compose.yml https://raw.githubusercontent.com/lfnovo/open-notebook/main/docker-compose.yml
  2. Step 2 — Generate an encryption key

    Make a long random secret for OPEN_NOTEBOOK_ENCRYPTION_KEY.

    openssl rand -hex 32
  3. Step 3 — Start the services

    Bring up SurrealDB and the app, then open the UI.

    docker compose up -d
// cas d'usage

Ce que vous pouvez créer

  • Research with cited answers from your own documents
  • Turn papers and notes into multi-speaker podcasts
  • Self-host a private NotebookLM for a team
  • Build research workflows on the REST API

Two more from the same shelf: insights-lm-public — Self-hosted NotebookLM alternative, and SurfSense — Air-gapped, open-source NotebookLM alternative that. Browse every RAG & Knowledge tool in the catalogue to compare the category. Before running a model locally, check how much VRAM you need for local LLMs — model size has to match your hardware.

// faq

Questions fréquentes

Is Open Notebook really free?

Yes — MIT licensed and self-hosted. You pay only for the AI provider usage you choose, or nothing with local Ollama models.

How is it different from NotebookLM?

It is self-hosted, works with any provider (not just Google), has a full REST API, and makes podcasts with up to 4 speakers.

Do I need Docker?

Docker Compose is the recommended route; a from-source install exists for developers.

Which models can I use?

20+ providers including OpenAI, Anthropic, Google, Mistral and Ollama, plus LM Studio and any OpenAI-compatible server.

Is my data private?

Yes — it is self-hosted, and with local models nothing leaves your machine. Note the UI has no password by default on a shared network, so bind it to localhost or set OPEN_NOTEBOOK_PASSWORD.

// continuez l'exploration

Outils similaires

RAG et connaissances

insights-lm-public

Self-hosted NotebookLM alternative — chat with documents, get audio summaries.

★ 656 étoilesMIT
Voir le guide →
RAG et connaissances

SurfSense

Air-gapped, open-source NotebookLM alternative that turns local documents into briefings, decks and podcasts.

★ 16,3 k étoilesApache-2.0
Voir le guide →
RAG et connaissances

deep-searcher

DeepSearcher — open-source deep research over your private data, with reasoning models and Milvus retrieval.

★ 8,3 k étoilesApache-2.0
Voir le guide →
RAG et connaissances

reader-source

Self-hosted rag & knowledge: open-source native macos ai reading workspace for pdf.

★ 58 étoilesMIT
Voir le guide →
RAG et connaissances

uteke

Self-hosted rag & knowledge: 🧠 the brain for your ai — local-first memory engine for.

★ 268 étoilesApache-2.0
Voir le guide →
RAG et connaissances

OpenDocuments

Self-hosted RAG search across GitHub, Drive, Notion, and local files.

★ 115 étoilesMIT
Voir le guide →
// alternatives

Alternatives en RAG et connaissances

Les autres projets de la catégorie RAG et connaissances avec le plus d'étoiles du catalogue LibreAI — la sélection réaliste d'alternatives si Open Notebook ne vous convient pas.

  • MarkItDown★ 193,5 k

    MarkItDown — Microsoft tool that converts PDFs, Office files, images and more to clean Markdown for LLMs, MIT

  • Dify★ 157,8 k

    Build Agentic workflows, RAG pipelines, with rich AI model and tool support on one collaborative workspace. Deploy on cloud, VPC, or…

  • ragflow★ 91,6 k

    RAGFlow is a leading open-source Retrieval-Augmented Generation (RAG) engine that fuses cutting-edge RAG with Agent capabilities to create a superior context…

  • PaddleOCR★ 90,8 k

    Lightweight OCR toolkit turning PDFs and images into structured, LLM-ready data in 100+ languages

  • MinerU★ 82,5 k

    MinerU — turns PDFs and Office documents into LLM-ready Markdown and JSON for agents and RAG pipelines, Apache-2.0 plus terms

// plus en rag

Plus en RAG et connaissances

Projets de la catégorie RAG et connaissances proches dans le catalogue, les plus proches en étoiles de Open Notebook — continuez à explorer de part et d'autre.

  • private-gpt★ 57,6 k

    Complete API layer for private AI apps on local models: RAG, skills, tools, MCP and text-to-SQL

  • faiss★ 41,1 k

    Meta's library for efficient similarity search and clustering of dense vectors, in C++ and Python

  • serviette★ 58,7 k

    A no-code RAG server that keeps answers in sync with your documents. Works with 8 vector databases. Built on Pathway.

  • Marker★ 40,6 k

    Marker — converts PDFs to Markdown and JSON quickly and accurately, with layout, tables and math intact, Apache-2.0

  • anything-llm★ 66,7 k

    Stop renting your intelligence. Own it with AnythingLLM. Everything you need for a powerful local-first agent experience

  • quivr★ 39,6 k

    Open-source engine turning continuous content streams into search, monitoring and alerts

Obtenir Open Notebook

Gratuit et ouvert : récupérez le code source, mettez une étoile au dépôt et commencez à créer.