What sources can NotebookLM use?
Natively: PDFs, plain text, Markdown, Google Docs, Google Slides, web URLs, YouTube videos (transcript extracted), audio files (auto-transcribed), and pasted text. With Notebook Toolkit, you also get one-click capture from ChatGPT, Claude, Gemini, Perplexity, YouTube, Reddit, LinkedIn, X, Substack, arXiv, PubMed, GitHub, and 20+ more platforms.
In more detail
NotebookLM's native source types are PDFs (up to 200 MB / 500K words), plain text and Markdown files, Google Docs and Slides, web URLs (NotebookLM fetches the page), YouTube videos (auto-transcribed), audio files (auto-transcribed), and pasted text. Per-source word limit: 500,000. Per-notebook: 300 sources (free) or 1,500 (Plus). Not natively supported: ePub, Word .docx (convert to PDF first), images (extract text via OCR first), tweets, Reddit threads, ChatGPT conversations. Notebook Toolkit fills these gaps with one-click capture from AI chats (ChatGPT, Claude, Gemini, Perplexity, DeepSeek, Mistral, Copilot), video (YouTube), social (Reddit, X, LinkedIn, Hacker News), academic (arXiv, PubMed), reference (Wikipedia), code (GitHub, Stack Overflow), and any web page.
Get one-click capture for NotebookLM
Notebook Toolkit saves ChatGPT, Claude, YouTube, Reddit, and any web page to NotebookLM in one click. Free to install.
Install Notebook Toolkit