Podcast to Transcript

Yes, when the episode is reachable at a direct audio address. The link is followed, the audio is retrieved, and the transcript is produced from it. An episode behind a private feed or a paywall cannot be fetched, so the downloaded file is used in that case.

Speaker separation is switched on before transcription. Each change of voice starts a new turn with its own label, so an exchange between the hosts and a guest stays readable. The labels are numbers until they are renamed, and renaming one label updates every turn that carries it.

Music at the start and end of an episode is transcribed as noise or skipped, which can leave a stray line in the first paragraph. A quick deletion fixes it. Music beds under speech are harder: they reduce accuracy slightly, though the words usually survive when the voice is mixed clearly.

Episodes of an hour or more are accepted, and the transcript covers the whole recording. Processing time grows with length, and the daily free allowance is measured in audio time. For a long interview show, transcribing one episode at a time keeps the files easy to identify.

It can, as a source the host reads and condenses. The transcript gives exact wording for quotes, guest names, and the order in which topics came up, and those passages can be lifted into notes or an article. Writing the condensed show notes itself remains a manual step.

The transcript follows what was said, so repeated words and half-finished sentences appear as they were spoken. Some light punctuation is added for readability. Removing every filler is closer to editing than transcription, and the text can be cleaned in the browser before export.

Accuracy depends on recording quality more than on accent. A close microphone and a quiet room produce a strong result across most accents; a remote guest on a poor connection is harder. Names, places, and specialist vocabulary are the terms most worth checking.

Yes. Plain text and a formatted document are both available, and a timed subtitle file can be produced from the same job when the episode appears on video. The text can also be copied paragraph by paragraph into a content system or a notes app, which is usually faster than building a script from a second listen.