YouTube Transcript to Podcast

Overview

Some videos contain material you want to hear without keeping the video open, but captions alone are not written for listening. This workflow gives you a podcast script and a synthesized MP3 from one selected YouTube video. It is the audio option, not the concise written-summary workflow. Get YouTube Transcript retrieves the available captions, Write Podcast Script reshapes them for spoken delivery, and Speak Podcast Script renders the script as audio. You can inspect and adapt this three-stage process with visual workflows.

Pipeline

1. Get YouTube Transcript receives one YouTube video URL or ID and returns its available captions.
2. Write Podcast Script receives the transcript and returns a script prepared for spoken delivery.
3. Speak Podcast Script receives the script and returns the synthesized MP3.

Inputs

• Source video. A YouTube video URL or ID. The default scope is one video per run.
• Script instructions. The guidance in Write Podcast Script. Keep the built-in podcast-script instructions for the first run, then edit them when you need a different structure or tone.
• Speech settings. The voice and language configured in Speak Podcast Script. Keep the node's current settings as the default unless the output needs a different voice or language.

Outputs

• Podcast script. A Markdown or text artifact containing the spoken copy.
• Podcast audio. An MP3 synthesized from that script.

Integrations

• Get Transcript. Reads the available caption track for the source video.
• AI. Rewrites the transcript as a podcast script.
• Speak. Synthesizes the script into an MP3.
• Visual workflows. Let you inspect stage handoffs and edit each node.

Sample output

The artifacts below are illustrative. They show the script and MP3 output shape, not a completed run.

podcast-script.md

# A practical guide to smaller releases
 
## Opening
 
Today we are looking at why small releases make software changes easier to review, observe, and reverse.
 
## Segment one
 
The first principle is scope. A narrow release gives the team fewer variables to inspect when behavior changes.
 
## Closing
 
Start with one reversible change, observe the result, and carry that evidence into the next release.

artifacts.txt

podcast-script.md
podcast.mp3

The MP3 is a binary artifact, so the illustrative sample lists its filename alongside the script rather than printing its contents.

FAQ

Does this use the video's original audio?

No. The workflow reads the available captions, writes a new podcast script, and synthesizes that script with Speak Podcast Script.

Can a missing caption track stop the MP3?

A missing caption track stops the pipeline at transcript retrieval. The script and speech stages need transcript text from Get YouTube Transcript.

Is the script stage necessary before speech?

Captions preserve transcript text, while the Write Podcast Script node prepares that text for spoken delivery. Speak Podcast Script generates the MP3 from the resulting script rather than the caption track.

Can I change the voice or language?

Yes. Edit the settings in Speak Podcast Script before the run. Keep the node's current settings when you want the configured default.

Related workflows

• YouTube Transcript Summarizer when you need a written Markdown and HTML summary instead of a script and MP3.
• RSS News Podcast when your source is a set of RSS feeds instead of one YouTube video.
• Daily AI News Digest when you need a written feed digest instead of synthesized podcast audio.