README

Purpose

Research on Google’s NotebookLM API capabilities for programmatically generating podcasts, audio overviews, and managing content automation. This document covers both the official Enterprise API and alternative solutions.

Key Findings

Official API Support (Enterprise Only)

Yes, Google provides an API - but only for NotebookLM Enterprise users with Gemini Enterprise licenses.

Podcast API

Method: podcasts - standalone API for generating podcasts
Audio Overview Method: notebooks.audioOverviews.create - generates audio overview for notebook content
Access: Requires IAM role roles/discoveryengine.podcastApiUser
Input: Array of multimedia objects (text, images, audio, video clips)
Customization:
- FOCUS prompt to guide podcast direction
- Language specification via BCP47 language tags
- Defaults to English if no language specified
Output: MP3 file downloadable via curl command

Key Differences

Podcast API: Standalone, doesn’t require a NotebookLM notebook or data store
Audio Overview: Must be associated with an existing notebook with data sources

Consumer Version Limitations

No public API for the free/consumer version of NotebookLM. Web interface only with these features:

Supports: PDFs, .txt files, Google Docs, Google Slides, web articles
Does NOT support: Spreadsheets, highly visual content
Can generate podcasts from Google Slides content
One-click audio generation from web UI only

Audio Overview Feature

What it does:

Transforms documents into engaging audio discussions
Two AI hosts conduct a “deep dive” discussion
Summarizes material, makes connections between topics
Provides objective reflection of source content
Available in 50+ languages (as of 2025)
Designed for use cases like course readings, legal briefs, research papers

Video Overview Feature (Announced I/O 2025, Launched July 2025)

This is the “Audio + Slides” feature - combines narrated audio with auto-generated animated slides.

What it does:

Transforms notebook sources into AI-narrated animated short films (5-15 minutes)
Analyzes up to 50 sources (500,000 words per source)
Generates cartoon-style visual effects, dynamic text, and AI narration
Uses Gemini’s Nano Banana image generation model for visuals

Format Types:

Explainer: Structured, comprehensive video for in-depth understanding
Brief: Bite-sized format to quickly grasp core ideas

Visual Styles (for 18+ users):

Classic, Whiteboard, Watercolor, Retro Print, Heritage, Paper-craft, Kawaii, Anime

Access via NotebookLM Studio panel - four tiles:

Audio Overviews
Video Overviews
Mind Maps
Reports

API Status for Video Overviews:

⚠️ No dedicated API found yet for Video Overviews
Currently UI-only via Studio panel
Enterprise Audio Overview API exists, but Video Overview API not documented
May be added in the future given enterprise demand

Supports multi-tasking: listen to Audio Overview while exploring Mind Map or reviewing Study Guide.

Slides Generator Feature (In Development - Oct 2025)

Upcoming feature - standalone slides generation (not video format):

Will integrate as a standard option in NotebookLM Studio panel
Will use Google’s Nano Banana model for image generation
Create presentations directly from sources without leaving NotebookLM
Part of Google’s strategy to make NotebookLM an “all-in-one knowledge to content pipeline”

Custom Automation Solutions

Developers have built custom implementations using Google Cloud tools:

Example Architecture:

Text processed by Gemini 1.5 Flash
Generates structured podcast script
Google TTS API converts script to audio
Volume Mounts for processing
Final podcast stored in Cloud Storage bucket

Third-Party API Alternatives

AutoContent API

NotebookLM alternative for developers
Multilanguage support
Custom voices
Advanced podcast controls
FAQ generation and study guides
Pricing: 10 credits per podcast (standard voices)
Input: Websites, YouTube videos, PDFs, plain text

Podcastfy (Open Source)

Python package for podcast generation
Multi-modal content support (text, images)
Multi-lingual audio conversations
Input: Websites, PDFs, images, YouTube videos
Focus on programmatic and customizable generation
Unlike NotebookLM: designed for scale and automation

Summary Table

Solution	API Access	Cost	Customization	Use Case
NotebookLM Enterprise	Yes (Podcast API)	Requires Gemini Enterprise license	Focus prompts, language selection	Enterprise automation
NotebookLM Consumer	No	Free	Limited (web UI only)	Personal research, manual generation
AutoContent API	Yes	Pay per podcast	High (custom voices, controls)	Professional content creation
Podcastfy	Open source	Free (self-hosted)	Very high (Python-based)	Developers, custom workflows

Answer to Original Question

Does Google provide an API to manage NotebookLM and trigger podcast and slide creations?

Yes, but with caveats:

Feature	Enterprise API	Consumer Access
Audio Overview (Podcast)	✅ `notebooks.audioOverviews.create`	Web UI only
Standalone Podcast	✅ `podcasts` method	Web UI only
Video Overview (Audio+Slides)	⚠️ Unknown (likely coming)	Web UI only
Slides Generator	⚠️ In development	Not available yet

Summary:

✅ Enterprise users: Full API access for podcasts via Podcast API
✅ Video Overview exists: “Audio + Slides” is the Video Overview feature (launched Aug 2025)
⏳ Standalone Slides: Coming soon as separate feature
❌ Consumer API: No programmatic access, web UI only
✅ Alternatives: Third-party APIs and open-source tools available

Sources

../google-cloud/ (if exists)
../ai-automation/ (if exists)