Gemini Enterprise & NotebookLM API Explorer

This application demonstrates the feasibility of interacting with Gemini Enterprise (previously known as Agentspace) and NotebookLM Enterprise functionality via API. This is helpful to prove that subscribers can integrate these capabilities into their own applications, services, and UI.

Use Cases

Gemini Enterprise: Use as an agent hub and integrate agents into your own app, without redirecting users to the Gemini Enterprise interface.

NotebookLM Enterprise: Create and manage notebooks programmatically, add data sources (text, web, YouTube, Google Drive, files), and share with team members.

Note: There is currently no programmatic support for querying (e.g. ask questions) NotebookLM.

Note: This explorer app demonstrates API integration using Application Default Credentials (ADC), not user OAuth credentials. For production applications serving end users, you would typically implement OAuth 2.0 for user authentication.

Prerequisites

Before running the application, ensure you have:

Python 3.9+
Node.js 18+ and npm
Google Cloud Project with Discovery Engine API enabled (The project where Gemini Enterprise is enabled). Ensure you have license and access to that Gemini Enterprise instance
Application Default Credentials (ADC) configured

Setting up Application Default Credentials

This application requires ADC to authenticate with Google Cloud APIs:

# Install Google Cloud SDK (if not already installed)
brew install google-cloud-sdk  # macOS
# Or download from: https://cloud.google.com/sdk/docs/install

# Authenticate and set up ADC
gcloud auth application-default login

# For NotebookLM with Google Drive access (optional)
gcloud auth login --enable-gdrive-access

# Verify authentication
gcloud auth application-default print-access-token

Important: Ensure you have the necessary permissions in your Google Cloud project to access Discovery Engine APIs and NotebookLM Enterprise.

🚀 Quick Start

git clone https://github.com/0nri/gemini-enterprise-api-explorer.git
cd gemini-enterprise-api-explorer
chmod +x setup.sh
./setup.sh

The script will:

✅ Check prerequisites (Python 3.9+, Node.js 18+)
✅ Create Python virtual environment
✅ Install all dependencies
✅ Start both backend and frontend servers
✅ Open your browser to http://localhost:3000

Then configure your credentials in the sidebar:

Enter your Google Cloud Project Number (not project ID)
Enter your Agentspace Engine ID
Click "Apply Configuration"
Start exploring!

Finding Your Configuration Values

Project Number: Google Cloud Console → Project Settings → Project number
Engine ID: Discovery Engine → Engines → Your engine name
Location: Usually us, eu, or global (default: us)

Features