Apple Podcasts Transcripts: Automatic Ones and Your Own File

Automatic transcripts

Apple introduced transcripts with iOS 17.4. For new episodes in supported languages, Apple creates a transcript automatically after publishing, and listeners open it from the episode's player. According to Apple's podcaster guidance, the supported languages are English, Danish, Dutch, Finnish, French, German, Italian, Norwegian, Portuguese, Spanish and Swedish. Apple notes some exceptions: very long episodes (over 10 hours) may not be transcribed, and song lyrics in music aren't shown.

Listeners see the text highlighted as the episode plays, can tap a line to jump there and can search inside the episode. For many shows, that's all you need.

Providing your own transcript

You might want your own transcript instead of Apple's:

• Your show uses names, jargon or a second language that automatic transcripts get wrong.
• You've already paid for a carefully proofread transcript and want listeners to see that version.
• You want speaker names shown.

Apple accepts transcript files linked from your RSS feed with the podcast transcript tag from the Podcasting 2.0 namespace, in VTT or SRT format. Most hosting providers add the tag for you when you upload a transcript file with an episode; check your host's episode settings for a transcript field. In VTT, speaker names can be marked with voice tags, so listeners see who is talking.

Making the transcript file with mydubly

To produce a VTT or SRT for an episode:

1. Export the final episode audio, the same file you publish, so the timings match exactly.
2. Upload it to audio to text and choose the language spoken.
3. Download the VTT or SRT along with the plain transcript.
4. Proofread names and terms. If you want speaker names, add voice tags in the VTT while you proofread; mydubly transcripts don't label speakers.
5. Attach the file to the episode in your hosting dashboard.

A 45-minute episode costs 45 credits, about 4.5 cents; new accounts start with 100 free credits. How to edit an SRT file covers fixing lines without breaking timings, and what a VTT file is explains the format, including voice tags.

Shows in languages Apple doesn't transcribe

If your show is in Japanese, Hindi, Korean, Arabic or another language outside Apple's list, Apple won't generate a transcript. Your own file through RSS is the route to a transcript in those cases, subject to Apple's handling of provided transcripts. Alongside the in-app transcript, many shows also publish the text on their website, which helps search engines and listeners who prefer reading. Publishing transcripts on your website covers layout and search benefits, and podcast show notes from a transcript covers turning the text into show notes.

Translated transcripts for a wider audience

A transcript in a second language can reach listeners who read it better than they understand the audio. From one upload, choose a target language among 21 and mydubly returns a translated transcript and subtitles. That translated text belongs on your website or in show notes rather than in Apple's transcript slot, which should match the spoken language. If you go further and publish a dubbed edition of your show, the multilingual podcast guide covers feeds and naming.

Limitations and quality checks

• Apple decides how provided transcripts are used and displayed; check its current podcaster guidance before relying on your own file.
• Your transcript must match the published audio. Re-export the file if you edit the episode after transcribing.
• Automatic transcripts struggle with crosstalk; two hosts talking over each other reads badly whichever tool you use.
• mydubly works from audio files up to 2 hours, so very long episodes need splitting and the files need merging.

Why transcripts matter for your show

Transcripts help deaf and hard-of-hearing listeners, people listening in a second language and anyone looking for one quote in a long episode. They also give you material for blog posts, newsletters and social clips. The podcast transcription page covers the recurring workflow for weekly shows.

A routine for a weekly show

Fold the transcript into the steps you already follow. After the final mix, export the episode audio, upload it while you write the episode title and description, and proofread the transcript while the episode uploads to your host. Keep a running list of guest names, recurring segment names and jargon, so proofreading takes minutes rather than a full listen. Store the VTT next to the episode audio with the same file name, so you can find it if you ever re-upload the episode. Shows with a back catalogue can work backwards a few episodes a week, starting with the most-played ones, rather than trying to transcribe everything at once; searching AI transcripts explains how a full set of transcripts becomes a searchable archive of everything said on the show.