Importing a NotebookLM Audio Overview
Turn a NotebookLM audio overview (or any audio file) into an editable Jellypod episode voiced by your own hosts.
Overview
If you have a finished audio overview from NotebookLM, you can import it into Jellypod and rebuild it as an editable episode in your own hosts' voices. Jellypod transcribes the audio, separates the speakers, and creates a draft episode with a full script you can edit before generating audio.
This is different from Uploading Your Own Episode, which publishes a recording as-is. Import from NotebookLM turns the audio back into a script you control, so you can rewrite lines, fix names, and regenerate it with your hosts.
How to Import
- Go to Podcasts in the sidebar and open the podcast you want to add an episode to.
- Click the upload icon button next to Create New Episode. It opens a menu with Upload Episode Script, Upload Episode Audio, and Import from NotebookLM.
- Select Import from NotebookLM.
- Pick your First Host and Second Host. These default to the podcast's regular cast, and you can swap either one. The two hosts must be different.
- Drag and drop your audio file into the upload area, or click Browse Files.
- Click Import Audio. A progress bar shows the status while Jellypod transcribes and imports.
- When it finishes, you land in the script editor with a new draft episode.
What Happens on Import
- Transcription and diarization: Jellypod transcribes the audio and separates it into speakers.
- Speaker mapping: the detected speakers are mapped onto the two hosts you chose, in the order they first appear. No new hosts are created.
- An editable draft: the result is a draft episode with a script, not a published recording. Nothing goes live until you generate audio and publish.
Once the draft is ready, edit it like any other episode: rewrite lines, reassign speech blocks, and regenerate individual segments in the Script Editor, then generate the episode audio in your hosts' voices.
Supported File Formats
- MP3: recommended for most use cases
- WAV: lossless audio, larger file size
- M4A: Apple's audio format
Imports are limited to 30 minutes of audio. If your file is longer, trim it before importing.
Most NotebookLM Audio Overviews have two speakers, which maps cleanly onto the two hosts you pick. If the audio has more than two voices, the extra speakers cycle back onto your two chosen hosts.
Non-speech sounds in the source audio, like [sighs] or [laughter], are ignored on import so they never turn into an extra speaker or stray line. To direct delivery in the rebuilt episode, add audio tags in the script editor.
Import from NotebookLM vs. the Other Upload Options
- Import from NotebookLM: transcribes audio, rebuilds it as an editable script, and voices it with your hosts. Creates a draft.
- Upload Episode Audio: publishes your recording as-is with no script or editing. See Uploading Your Own Episode.
- Upload Episode Script: imports a written script (not audio) and voices it with your hosts. See Importing and Exporting Scripts.
Frequently Asked Questions
Does importing from NotebookLM publish the episode?
No. It creates a draft episode with an editable script. Nothing is published until you generate the audio in your hosts' voices and publish it yourself.
Will it create new hosts for the speakers?
No. The detected speakers are mapped onto the two hosts you pick (defaulting to the podcast's regular cast), in the order they first appear. No new hosts are created.
What if the audio has more than two speakers?
Imports use exactly two hosts. If the audio has more than two voices, the extra speakers cycle back onto your two chosen hosts. You can reassign any speech block to a different host in the script editor afterward.
Is there a length limit?
Yes. Imports are limited to 30 minutes of audio. Trim longer files before importing. Accepted formats are MP3, WAV, and M4A.
Can I import audio that isn't from NotebookLM?
Yes. Any MP3, WAV, or M4A file works. NotebookLM is just the most common starting point.
Was this page helpful?