Podcast Transcript Generator

Before you look for a tool, check whether the app you listen in already has a transcript. Two of the biggest podcast apps now show them for many episodes.

Apple Podcasts. Apple added transcripts in iOS 17.4 and iPadOS 17.4. Apple says it makes them automatically after a new episode is published, and that creators can supply their own instead. On iPhone or iPad, play the episode, open the player and tap the transcript button. A greyed-out button means that episode has none. You can search the transcript, tap a paragraph to jump to that point, and select and copy up to 200 words at a time to share. Apple notes a few gaps: transcripts can take a while to appear after an episode goes live (it suggests waiting 24 hours), music lyrics are not shown, and episodes outside the Apple Podcasts catalog do not get them.

Spotify. Spotify's help page says some shows have auto-generated transcripts. Listeners see them in the Now Playing view and on episode pages in the Spotify mobile app. Not every show has them, and the page does not describe a way for listeners to download the text.

YouTube. If the episode is on YouTube, you can read it there too: open the description and select Show transcript. YouTube says this shows the full transcript of any video that has captions. Our tool reads the same captions and lets you download them as TXT, SRT or VTT, with a short summary.

Audio only, and you need a file you can edit. Then you need a different kind of tool. Some speech-to-text tools let you upload the audio file (MP3, WAV and so on). Others take a Spotify or Apple Podcasts episode link. Some podcast hosts also make transcripts for the shows they host. Check each tool's own page for what links and files it accepts and what it charges.

Because this tool reads YouTube captions, the quality of your transcript is the quality of those captions. Automatic captions come from YouTube's speech recognition, and YouTube warns they can get words wrong because of accents, dialects, mispronunciation or background noise. The show format changes how much cleanup that means.

Solo shows are the easy case: one voice and no crosstalk. The transcript often needs little more than paragraph breaks and a check of product and place names.

Interview shows add two problems: people talking at the same time, and names the captions may not know, like the guest, their company or a book title. Search the transcript for those names first and fix them. They are the words listeners will search for and quote. The tool does not add speaker labels, so mark who is talking where it matters.

Panel shows (three or more voices) give the roughest text. YouTube lists overlapping speakers as one reason automatic captions may not be made at all. Plan a real editing pass. For a panel, a cleaned-up "key moments" transcript is often more useful than every half-sentence.

Whatever the format, the most useful edit is the same: fix the names and keep the spoken style. If you polish talk into formal prose, quotes stop sounding like the people who said them.

The transcript is raw material. These three things are what you build from it:

1. Show notes. Skim the text and pull out the main topics with a one-line summary each, every book, tool or link someone mentions, and one strong quote. Searching the text for phrases like "check out" or "called" helps you find the mentions. The AI summary gives you a starting point, but check it against the transcript.
2. Chapters. Export the SRT or VTT file, which has timestamps, and note the lines where the topic changes. For YouTube, paste the list into the description. YouTube makes chapters when the first timestamp is 00:00, there are at least three timestamps in order, and each chapter is at least 10 seconds long. Use the same list in your podcast host's chapter editor if it has one.
3. The episode page. Post the cleaned-up transcript on the episode's page on your own site, under the player. Readers can then find a quote without listening again, and listeners who are deaf can follow the episode.

If you are the creator, you can also send the transcript back to the podcast apps. Spotify's creator help says eligible creators can upload a VTT or SRT file of up to 5MB, with timestamps (turning on Spotify's auto-generated transcripts replaces it). Apple reads creator transcripts from the RSS feed through the transcript tag, so your host has to support it. Only do this when the YouTube video and the audio episode are the same edit. If the video has a different intro or ad breaks, the timestamps will be off.