Use case

Podcast Translator

Publish your show in another language. Each episode becomes a translated M4A you can release in a separate feed, along with transcripts in both languages for your show notes.

01 Upload

Your audio

Audio and text outputTranslated audio, playback, and transcripts in the result panel.
Server assistOff by default. When on, this job uploads your media file so our servers can decode it (helpful on phones). Uploads this file (max 500 MB). Same credit cost as a normal run. Max length still 2 hours.

By default the audio file stays in this browser (only processed audio chunks go to the cloud). Language is required for translated audio (21 targets; voice in 19).

02 Result

Your translated audio

Your translated audio will appear here.

Waiting for upload

25 languages available

Why it helps

Podcast audiences are fragmented by language, and new-language feeds are a proven way to grow — but hiring voice actors for every episode is expensive. AI translation makes a pilot feed affordable.

Step by step

  1. Export the episode as MP3 or WAV.
  2. Choose the target language; each host and guest is voiced in a copy of their own voice, so close-mic recordings work best.
  3. Download the translated M4A — your intro music and stings stay in the mix — and publish to a separate feed.

Tips

  • Translate a back-catalogue favourite first and share it with a few native-speaking listeners for feedback.
  • Tell listeners the episode is AI-translated — transparency builds trust.

Example

41 minutes

A Spanish-language history podcast translates a 41-minute episode into English for 2,050 credits ($2.05) and launches an English feed.

Cost: 4,920 credits ($4.92).

Limitations to know

  • Hosts are told apart automatically (up to 8 people); crosstalk, quick interjections or two similar-sounding hosts can put a line in the wrong voice, and overlapping speech overlaps in the dub by at most about two seconds.
  • Music is kept under the translated voice, but separation isn't perfect: sung vocals in music are removed, and faint traces of the original voice can remain where music is loud.

Decide what the translated show will be

Separate feed
A new show in the target language with its own title, artwork and description. Clearest for listeners, and directories list it as a show in its own right.
Episodes in the main feed
Translated episodes added to your existing feed, with the language in the title. Quick to launch, but confusing for subscribers who don't speak that language.
Translated transcripts only
No audio: a transcript-mode run with a target language at 5 credits per minute, posted on the episode page. A low-cost way to test interest.

Shows that commit usually go for a separate feed; multilingual podcasts compares the setups in more depth.

Preparing the source audio

What you upload shapes what comes back. If you upload the full mix, its music beds and stingers are kept under the translated voices, though loud music under speech makes both recognition and separation harder. If your editor can export the conversation without the music, recognition has less to untangle, but you will then add the music to the translated episode yourself. Cut long cold opens built from clips of later moments, since they will be translated again when the clip appears in full. And if the episode is close to two hours, split it at a natural break, such as the halfway sponsor slot, and join the translated halves in your editor.

Finishing the episode around the translated voices

The translated M4A holds the translated voices with the original background mixed underneath, in mono. It is close to a finished episode, but a few steps remain.

  1. Import the M4A into your editor on its own track.
  2. Check the intro, outro and transition music. Sung lyrics are removed along with the speech, and in dense mixes traces of the original voice can remain; if you have clean music stems, dubbing a dialogue-only export and mixing the translated voices with them gives the cleanest result.
  3. Deal with sponsor reads: re-record them with a speaker of that language or drop them, depending on your agreement. See sponsored segments in translated content.
  4. Normalise loudness to the same target as your main feed so listeners don't reach for the volume; loudness normalisation explains the targets.
  5. Export at your usual podcast settings and upload to the new feed.

A short spoken intro from the host, recorded in the new language if they can manage a few lines, makes the translated show feel less anonymous.

Interview and panel shows

Each person in the episode is dubbed in a copy of their own voice, so host and guest stay distinct in the translated version and a show with several hosts keeps them apart. mydubly works out on its own who says which line, for up to 8 people, and that step can slip: two people talking over each other, quick interjections like "yeah" or "mm", or two hosts who sound very alike can put a line in the wrong voice. Overlapping speech comes out as separate lines, each voiced on its own. A guest who only says a few words is voiced with the copied voice of whoever they sound most like, or a stock voice matching theirs. If the format depends on fast banter, listen to a full translated episode before committing; single-voice vs multi-voice dubbing discusses the trade-offs. Choosing an AI voice covers how listeners react to synthetic voices. For a consistent sound from episode to episode, record everyone on a close microphone, since each person's voice is copied from a short, clean sample of their own speech.

Budget for a pilot season

Three-episode pilot

Episodes of 35, 42 and 50 minutes, 127 minutes in total, dubbed into German: 6,350 credits ($6.35).

Translated show notes

Download translation in each dub run gives the full translated text of the episode at no extra cost, ready to edit into notes.

Monthly after launch

Four 45-minute episodes a month: 9,000 credits ($9.00) per language.

Give the pilot a fixed run of several episodes and judge it on downloads and messages from listeners in that language, not on the first week. New feeds start with no subscribers, and directories take time to surface them, so an early quiet spell says little about long-term interest.

Telling listeners and gathering feedback

Say at the start of each translated episode, and in the show description, that the voices are AI-generated copies made from the original episode. Listeners are more forgiving of a synthetic voice when they know what they are getting. Ask a handful of native-speaking listeners to report anything that sounds wrong, such as names, numbers or idioms translated literally, and keep a list. Fixing a single word in finished audio is fiddly, so those notes mostly improve how you prepare the next episode: spelling out an unusual name on air, or avoiding a pun that won't survive translation. The audio dubbing tool page lists every download, and podcast transcription covers transcripts in the original language.

Using the tool on this page

  • The spoken language is detected automatically across about 100 languages; you can pin it, or pick one of 21 languages to translate into. Files can be up to 2 hours long.
  • The finished dub opens in the dub editor, where you can correct words, timing, levels, background and subtitles and re-process only what changed. Its files are deleted 120 minutes after you last use the editor (how files are protected).
  • You can download the translated audio (voice over your original music and effects) as M4A, a transcript and subtitles (SRT and VTT). Every output, step and limit is explained on the audio dubbing page.

Frequently asked questions

Does each host keep their own voice?

Yes. Each host and guest is voiced in a copy of their own voice, taken from a short, clean sample of their speech in the episode.

Can I translate only the transcript?

Yes — switch to text-only for a translated transcript at 5 credits per minute.

Do I need a separate RSS feed for the translated version?

Not strictly, but it is usually clearer. A separate feed gives the show a title and description in the new language and keeps existing subscribers from receiving episodes they can't understand.

What happens to the music and ads in the original episode?

The original speech is separated out and the music stays under the translated voices, lowered while they speak; sung lyrics are removed along with the speech. Ad reads recorded into the episode are translated like any other speech; ads your host inserts later aren't in the file at all.

Can a listener translate an episode for personal listening?

If they have the audio file saved on their device, yes; the audio dubbing tool works on files, not podcast links. Republishing someone else's show needs the owner's permission.