In this categoryGemini ยท 24
- How to Install the Gemini CLI and Run Your First PromptStart
- How to Get a Gemini API Key from Google AI Studio
- How to Call the Gemini API from Python
- How to Give Gemini CLI Project Context with a GEMINI.md File
- How to Choose the Right Gemini Model for Coding Tasks
- How to Fix a Bug in Your Codebase with the Gemini CLI
- How to Set Up Gemini Code Assist in VS Code
- How to Stream Gemini Responses in Node.js
- How to Feed a Long Document into Gemini's Long Context
- How to Run the Gemini CLI Non-Interactively in Scripts
- How to Analyze a Video File with the Gemini API
- How to Connect an MCP Server to the Gemini CLI
- How to Fix Common Gemini CLI Authentication Errors
- How to Cache Long Context in the Gemini API to Cut Costs
- How to Summarize a YouTube Video with Gemini
- How to Analyze PDFs and Docs in the Gemini Chat App
- How to Extract Data from Images and Screenshots with Gemini
- How to Generate a Short Video with Veo in Gemini
- How to Ground Gemini Answers in Your Own Documents
- How to Use Gemini Inside Google Docs to Draft Faster
How to Fix a NotebookLM Source That Will Not Load
Diagnose why a PDF, URL, or YouTube link fails to import into NotebookLM and apply the right workaround.
Sometimes a source in NotebookLM gets stuck importing, shows an error, or comes in empty. Usually the cause is a scanned PDF, a blocked URL, or a file that is too large. This guide runs through the common failures and the fix for each.
What you need
- The source that failed to import
- A way to convert or re-export files (a PDF viewer or word processor)
- A few minutes
Step 1: Identify the failure type
Look at what the source row shows. An error icon, a stuck spinner, or a source that imports but returns nothing each point to different causes. Note which one you are seeing.
Step 2: Match the symptom to a fix
Use the table below to jump to the right remedy. The most common culprit is a scanned PDF that has no selectable text layer for NotebookLM to read.
| Symptom | Likely cause | Fix |
|---|---|---|
| PDF imports but chat finds nothing | Scanned image, no text layer | Run OCR, then re-upload |
| URL import fails | Site blocks crawling or needs login | Copy the text and paste it as a source |
| YouTube link rejected | No captions or private video | Use a public video with captions |
| Upload stalls | File too large | Split the file or export a lighter PDF |
Step 3: OCR a scanned PDF
If your PDF is a scan, add a text layer with OCR before re-uploading. A free command-line tool like ocrmypdf does this in one step.
Step 4: Fall back to pasted text
When a URL refuses to import, open the page, select the article body, copy it, and use the Paste text source option. This bypasses any crawling restrictions because you are providing the content directly.
Result
After OCR or pasting the text, the source imports cleanly and the chat can cite it. Most NotebookLM loading problems trace back to scanned PDFs or blocked URLs, and both have quick workarounds.
Watch related tutorials
19:27
16:12
08:00
20:00
14:22
09:15New guides in your inbox
Fresh step-by-step how-to guides as we publish them. One email a week, no more.