Import NotebookLM audio as an editable episode

Import from NotebookLM turns a finished audio file into a draft episode with an editable script, voiced by two of your hosts. It works from any podcast in Podcasts and Episodes, and any audio file works, not only NotebookLM overviews. To publish a recording as-is instead, see which import option to use.

Import an audio file

1. Open Podcasts in the sidebar and select your podcast.
2. In Create New Episode, select Import from NotebookLM.
3. Choose a First Host and a Second Host. They default to the podcast's regular cast, and they must be different. The podcast needs at least two hosts.
4. Drag the audio file into the upload area, or click Browse Files.
5. Click Import Audio. Keep the dialog open while it uploads, transcribes and imports, because it cannot be closed mid-import.

When it finishes, the script editor opens on the new draft.

What import does

• Transcribes and separates speakers. Detected speakers map onto your two hosts in order of first appearance. No new hosts are created.
• More than two speakers. Extra speakers cycle back onto your two hosts. You can reassign any speech block in the script editor afterward.
• Ignores non-speech sounds. Sounds like [sighs] or [laughter] are dropped, so they never become a speaker or a stray line. To direct delivery, add audio tags after import.
• Starts audio generation. Narration begins right away in your hosts' voices and uses credits (see understanding credits). If you don't have enough, the script is still saved and you can generate audio from the editor later.
• Publishes nothing. The result is a draft until you publish it.

Edit the draft in the script editor: rewrite lines, fix names and regenerate segments.

Limits

Files over 30 minutes or 250 MB are rejected, so trim them first. Enterprise has no plan cap on episode length.

Troubleshooting

• "Audio is too long": the file is over 30 minutes. Trim it and try again.
• "File too large": the file is over 250 MB. Use a smaller file.
• "Please upload an audio file": the file is not audio. Use MP3, WAV or M4A.
• "Episode script truncated": the audio ran past your plan's episode length, so the script was cut and an upgrade prompt opens. The truncated draft is still created.
• "We couldn't detect any speech in this audio": the file has no recognizable speech. Try a different file.
• Hosts unavailable or fewer than two: add at least two hosts to the podcast, or pick hosts that are still active.

Open Help & Support in the Studio sidebar to chat with the team, or email support@jellypod.com.