NotebookLlaMa🦙

A fluffy and open-source alternative to NotebookLM!

https://github.com/user-attachments/assets/7e9cca45-8a4c-4dfa-98d2-2cef147422f2

A fully open-source alternative to NotebookLM, backed by LlamaCloud.

Prerequisites

This project uses uv to manage dependencies. Before you begin, make sure you have uv installed.

On macOS and Linux:

curl -LsSf https://astral.sh/uv/install.sh | sh

On Windows:

powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"

For more install options, see uv's official documentation.

---

Get it up and running!

1. Clone the Repository

git clone https://github.com/run-llama/notebookllama
cd notebookllama/

2. Install Dependencies

uv sync

3. Configure API Keys

First, create your .env file by renaming the example file:

mv .env.example .env

Next, open the .env file and add your API keys:

• OPENAI_API_KEY: find it on OpenAI Platform
• ELEVENLABS_API_KEY: find it on ElevenLabs Settings
• LLAMACLOUD_API_KEY: find it on LlamaCloud Dashboard

🌍 Regional Support: LlamaCloud operates in multiple regions. If you're using a European region, configure it in your .env file:

• For North America: This is the default region - no configuration necesary.
• For Europe (EU): Uncomment and set LLAMACLOUD_REGION="eu"

4. Activate the Virtual Environment

(on mac/unix)

source .venv/bin/activate

(on Windows):

.\.venv\Scripts\activate

5. Create LlamaCloud Agent & Pipeline

You will now execute two scripts to configure your backend agents and pipelines.

First, create the data extraction agent:

uv run tools/create_llama_extract_agent.py

Next, run the interactive setup wizard to configure your index pipeline.

⚡ Quick Start (Default OpenAI): For the fastest setup, select "With Default Settings" when prompted. This will automatically create a pipeline using OpenAI's text-embedding-3-small embedding model.

🧠 Advanced (Custom Embedding Models): To use a different embedding model, select "With Custom Settings" and follow the on-screen instructions.

Run the wizard with the following command:

uv run tools/create_llama_cloud_index.py

6. Launch Backend Services

This command will start the required Postgres and Jaeger containers.

docker compose up -d

7. Run the Application

First, run the MCP server:

uv run src/notebookllama/server.py

Then, in a new terminal window, launch the Streamlit app:

streamlit run src/notebookllama/Home.py

[!IMPORTANT]

You might need to install ffmpeg if you do not have it installed already

And start exploring the app at http://localhost:8501/.

---

Contributing

Contribute to this project following the guidelines.

License

This project is provided under an MIT License.