How to Fix NotebookLM Can't Import YouTube Video (Comments Disabled, Captions Missing)
AI transcription

How to Fix NotebookLM Can't Import YouTube Video (Comments Disabled, Captions Missing)

You've probably been here: you copy a YouTube URL, paste it into NotebookLM, hit "Add source," and instead of the smooth import you expected, you get hit with "This video cannot be imported. Transcript not available." The video plays fine…

May 17, 20268 min read

You drop a YouTube link into NotebookLM and get "transcript not available" even though you can see the video playing fine. The usual fixes don't work. Here's what's actually happening and the fastest workaround.

You've probably been here: you copy a YouTube URL, paste it into NotebookLM, hit "Add source," and instead of the smooth import you expected, you get hit with "This video cannot be imported. Transcript not available." The video plays fine on YouTube. Comments might be disabled, or captions might not be there, but the video exists and you can watch it. So what gives?

The short version: NotebookLM only works with YouTube videos that have existing captions (either auto-generated or manually added by the creator). If captions don't exist, if they're disabled, if comments are off and that somehow blocks the transcript, or if the video is still processing, NotebookLM can't help you. It doesn't generate transcripts. It only pulls what YouTube already has.

That limitation is by design, not a bug. But when you need notes from a video that NotebookLM won't import, design doesn't matter. You need a workaround.

Why NotebookLM won't import certain YouTube videos

Three main reasons surface over and over in support threads and Reddit posts.

No captions exist. Some creators explicitly disable auto-captions on their videos. Others upload content where YouTube's auto-caption system fails (heavy accents, technical jargon, music, background noise, multiple speakers talking over each other). If YouTube never generated captions, NotebookLM has nothing to pull.

Video is still processing. Fresh uploads sometimes show as available on YouTube but the caption track hasn't finished generating yet. NotebookLM tries to grab it, gets nothing, throws an error. Come back in an hour and it might work. Or it might not.

Restricted access or metadata issues. Age-restricted videos, region-locked content, unlisted videos with certain privacy settings, and videos with comments disabled can all fail import for reasons that aren't clearly documented. Google's official NotebookLM help page confirms that only public videos with captions work reliably.

None of these are fixable on your end if you're not the video owner. You can't force YouTube to generate captions. You can't change the creator's privacy settings. If NotebookLM says no, it means no.

The manual workaround (copy-paste the transcript)

If the video does have captions and you just want to bypass NotebookLM's finicky import, here's the workaround that actually works.

On desktop YouTube, open the video. Under the title, click the three-dot menu ("More actions") and select "Show transcript." A panel opens on the right with timestamped text. Click once inside the panel, hit Ctrl+A (Windows) or Cmd+A (Mac) to select all, then copy it.

Back in NotebookLM, instead of adding a YouTube source, click "Add source" and choose "Paste text." Paste the transcript. Done. NotebookLM treats it like any other text source. You can chat with it, generate study guides, build audio overviews, the whole workflow.

The downside: this only works if the video has captions in the first place. If the video doesn't have a transcript to copy, you're stuck. That brings us to the actual fix.

The real fix: use a tool with its own transcription engine

NotebookLM doesn't transcribe. It imports. If the import fails, you need something that generates the transcript itself instead of relying on YouTube's caption system.

This is where VidNotes earns its keep. It's built to handle exactly this scenario. You paste a YouTube link and it tries three fallback methods in sequence:

  1. Pull YouTube's existing captions (same as NotebookLM tries)
  2. Hit a secondary transcript API if captions aren't available
  3. Download the audio and run it through OpenAI's Whisper transcription model

That third layer is why VidNotes works on videos where NotebookLM fails. Even if the creator disabled captions, even if comments are off, even if it's a fresh upload still processing, Whisper generates a transcript directly from the audio. No captions required.

The YouTube to transcript tool is the one-field version if you just want the text. Paste the link, get the transcript, copy it wherever you need it (including back into NotebookLM as a text source if you want the multi-source notebook workflow NotebookLM does well).

For the full AI workflow (summary, flashcards, chat with the video, action items), use the main VidNotes app. Same paste-and-go input, but you get the whole suite of outputs instead of just the transcript. More on that in our guide to chatting with any video using AI.

Comparison table: NotebookLM vs VidNotes for YouTube imports

Here's how the two tools handle the same problem.

ScenarioNotebookLMVidNotes
Video with auto-generated captionsWorksWorks
Video with manual captionsWorksWorks
Video with no captions at allFails ("transcript not available")Works (Whisper fallback)
Comments disabledMay failWorks
Age-restricted videoFailsMay fail (depends on restrictions)
Region-locked videoFailsMay fail (depends on region)
Fresh upload still processingFailsWorks (Whisper fallback)
Time from paste to transcript~30 seconds if it works1-3 minutes average
Export transcript as fileLimitedYes (TXT, SRT, VTT, PDF)
Works on TikTok, Instagram, VimeoNoYes
Multi-source synthesisYes (key strength)No

NotebookLM wins when you're building a research notebook with 20+ sources. VidNotes wins when you need one YouTube video transcribed no matter what.

Other platforms that work when NotebookLM doesn't

If your video lives outside YouTube entirely, NotebookLM isn't an option. It only supports YouTube and Google Drive video files. VidNotes handles a wider range of sources:

For a broader look at alternatives, the VidNotes vs NotebookLM comparison page goes feature-by-feature.

FAQ

Why does NotebookLM work on some videos but not others? It depends entirely on whether YouTube's caption system generated a transcript for that video. Public videos with auto-captions or manual captions work. Everything else is a gamble. Age restrictions, privacy settings, disabled comments, missing captions, and fresh uploads all increase the chance of failure.

Can I force YouTube to generate captions for a video? Only if you're the video owner. As a viewer, you can't control whether captions exist. If the creator disabled auto-captions or YouTube's system couldn't generate them (non-speech audio, heavy music, multiple overlapping speakers), you're out of luck with caption-dependent tools.

Does the copy-paste workaround always work? Only if the video has a transcript to copy. If there's no "Show transcript" option in the three-dot menu, there's nothing to paste. You'd need a tool with its own transcription engine at that point.

How accurate is Whisper transcription compared to YouTube captions? On clean audio (single speaker, minimal background noise, standard accent), Whisper accuracy is in the high 90s. On messy audio (multiple speakers, music, technical jargon, heavy accents), it's still better than nothing and often beats YouTube's auto-captions. You'll want to skim the output either way.

Is VidNotes free? Free trial that covers full functionality (transcription, summaries, flashcards, AI chat). After that it's $9.99 per month or $49.99 per year. NotebookLM is free. If cost is the only constraint and your videos work with NotebookLM, stick with NotebookLM. If reliability matters more, the paid tier is worth it.

Can I use VidNotes to transcribe a video and then paste it into NotebookLM? Yes. Generate the transcript in VidNotes, export it as TXT, paste it into NotebookLM as a text source. You get the best of both tools: VidNotes handles the transcription, NotebookLM handles multi-source synthesis and audio overviews.

When to stick with NotebookLM anyway

Credit where it's due. NotebookLM does things VidNotes doesn't try to do.

If you're writing a research paper with 30 sources (PDFs, Google Docs, YouTube videos, meeting notes) and you need to ask questions across all of them with citations, NotebookLM is the right tool. The multi-source synthesis is the killer feature. The audio podcast overview is genuinely impressive. The mind maps help visualize relationships between concepts.

But if your workflow is "I have one YouTube video, give me a transcript and a summary," NotebookLM is overkill. And if the video fails import because captions don't exist, it's not just overkill. It's a dead end.

For that workflow, VidNotes is the faster path. Try it free, paste the link that NotebookLM rejected, and you'll have a transcript in under three minutes. Works on iOS, Android, web, and Chrome.

If you're hitting the "transcript not available" error on a video you know exists and plays fine, the problem isn't you. It's the limitation of caption-dependent tools. Use a tool that generates its own transcripts instead.

Related tool

Get a YouTube transcript instantly

Paste any YouTube link and get the full transcript with timestamps, AI summaries, and flashcards.

Open tool

Get started

Turn your next video into searchable text in under a minute

Try VidNotes free in your browser — 3 transcriptions per month, no account required.