Audio guide
How to download a podcast episode with yt-dlp
Most podcast episodes are just an audio file sitting at a URL, so yt-dlp downloads them well. The only thing that changes is where the episode lives: a YouTube video, a direct MP3 in an RSS feed, or a host like SoundCloud with its own extractor. All three come down to one command with a couple of flags.
Quick answer
# Episode published on YouTube -> MP3
yt-dlp -x --audio-format mp3 "https://www.youtube.com/watch?v=VIDEO_ID"
# Direct episode file (what an RSS feed points at)
yt-dlp "https://traffic.example.com/episodes/ep-42.mp3"-x extracts the audio and --audio-format mp3 converts it (that step needs ffmpeg). A direct .mp3 URL is already audio, so you can drop -x and just download it.
Podcasts published on YouTube
A lot of shows post full episodes to YouTube. Pass the watch URL and pull the audio only so you are not downloading a video you will never watch:
yt-dlp -x --audio-format mp3 "https://www.youtube.com/watch?v=VIDEO_ID"
# [youtube] Downloading 1 format(s): 251
# [ExtractAudio] Destination: <title>.mp3Want chapter markers kept, or the MP3 tagged with the episode title and date? Add --embed-metadata --embed-chapters. For quality, the extract-audio guide covers --audio-quality and when to skip the MP3 conversion entirely.
Direct files and RSS feeds
Every podcast RSS feed points each episode at a plain media URL (the <enclosure> tag). Open the feed, copy the episode's .mp3 (or .m4a) URL, and hand it to yt-dlp — its generic extractor downloads it directly:
yt-dlp -o "%(title)s.%(ext)s" "https://traffic.example.com/episodes/ep-42.mp3"
# [generic] Extracting URL
# [download] Destination: ep-42.mp3This is also why most Apple Podcasts episodes download fine: the Apple page is just a listing on top of the same public RSS enclosure, so the direct-file method works once you have the real episode URL.
SoundCloud and other hosts
yt-dlp ships dedicated extractors for a lot of podcast and audio hosts, so for those you can paste the normal page URL instead of hunting for the file. Check whether yours is covered:
yt-dlp --list-extractors | grep -iE 'podcast|acast|soundcloud|megaphone|iheart'
# acast
# iheartradio:podcast
# megaphone.fm
# soundcloudFor a host on that list, yt-dlp "<episode page URL>" is enough. If yours is not listed, fall back to the direct RSS-enclosure URL above.
Grab a whole show at once
Point yt-dlp at a SoundCloud user, a YouTube playlist, or a channel and it downloads every episode. Number them so they sort in order:
yt-dlp -x --audio-format mp3 \
-o "%(playlist_index)02d-%(title)s.%(ext)s" \
"https://soundcloud.com/<show>/sets/<season>"Use --download-archive archive.txt so re-running only pulls new episodes. The playlists guide and the output-template guide go deeper on naming and archiving.
What yt-dlp can't do here
Spotify-exclusive and other DRM-protected shows can't be downloaded — the audio is encrypted, and that is a hard stop, not a flag you are missing. yt-dlp works on episodes that are served as a normal audio file somewhere public, which is still the large majority of podcasts.
Turning the episode into text
Podcasts almost never ship captions, so unlike downloading subtitles there is usually nothing to pull. The free route is to download the audio, then run a local speech-to-text tool like Whisper over the MP3 to get a transcript. It works, but transcribing a backlog of hour-long episodes on your own machine is slow.
Recurring sources
Pulling from the same sources on repeat?
Paste a link to preview what a maintained import API returns for it — across 500+ platforms.
Running it on a server?
Pulling episodes on a schedule from a cloud box is where this gets fiddly: datacenter IPs draw more checks on the hosted platforms, and a local Whisper run doesn't fit neatly in a cron job or a function. If you are hitting that, the yt-dlp blocked on a server guide covers the options.