How to Transcribe a Podcast in Descript

Most “transcribe a podcast in Descript” posts treat the script as finished the moment the spinner stops. It is not. You still have a guest named “Sara” who is Sarah, a product that came out as “sync the size,” and a choice: leave with a text file, or stay in that script and cut the episode. This guide is the first job — import, wait, names, export or edit — using the one editor we send for this brief: Descript.

When a transcript is the job — and when you should cut, clean, or caption instead

Use this pass when the words are the deliverable. A weekly interview that needs show notes, a guest who asked to review quotes, a producer who will not open the editor, a blog post that should match the episode: you import the file, you wait, you label the people, you Correct the names, you export text. The same script is also the cut surface if the next job is the radio edit. That is a later page. This one stops at usable words.

Skip this pass if you already trust the script and the remaining job is the ramble. Deleting the tangent at 18:40 is How to edit a podcast in Descript — an ordinary guide link, not a second money hop. Skip both if the file still sounds like a kitchen: How to clean up podcast audio in Descript. Long gaps are How to remove silence in Descript. Type on the picture is How to add captions in Descript. Those pages hop to Descript too. This one already will.

The seven-step Descript transcription pass

1. 01

   Confirm the episode is yours, then get a real audio file — not an RSS dump

   This pass is for a show you host, or a session you have written permission to transcribe and quote. Guest likeness, a licensed bed, and a sponsor read are not automatically yours to publish as text. Get the quote and show-notes rights in the guest release. Then export a recording: a mixed WAV/MP3 from the booth, a video podcast MP4, host and guest tracks from a recorder, or a Rooms session you captured in Descript. An RSS XML feed is not a transcription input. A published enclosure you already posted is a last resort — you would be transcribing a compressed copy.

2. 02

   Start a project for this episode and import the recording into the script

   Treat one episode as one project. Official getting-started copy gives you two starts: record in Descript, or import an existing file. Import paths they publish include upload from your computer, import from Zoom, and upload from a phone. Drag the file into the Script editor, or insert it from the Project panel. Official add-a-file help is blunt: only script media is transcribed. A file sitting as a layer is not a transcript yet. You can also run Transcribe file from the options menu next to the file name. Every uploaded or recorded file counts against media minutes.

3. 03

   Wait for the script — do not correct names while the file is still processing

   Transcription starts after the file has finished processing and reached the cloud. Official automatic-transcription help: time depends on length, audio quality, and your connection. Files longer than 15 hours fail to transcribe automatically; split those first. Music and song lyrics are not supported — Descript transcribes speech. If you work in more than one language, turn on Always ask before transcribing or set Default transcription language in App Settings before you import. Exiting the language picker leaves the file untranscribed. Wait until the script is there. Then read the open.

4. 04

   Identify speakers once — host and guest labels follow the rest of the episode

   Automatic speaker detection is on by default. Official Detect and label speakers help: when detection finishes you get a notification; click Identify speakers, name each voice or pick an existing label, then Close. Add the file to the script if it is not already there — the labels appear in the document. If detection was off, open the file options menu in the Project panel and choose Detect speakers. After that, official Speakers help: type @ in the script, or select text and Change speaker, then create or rename the label. Rename from any instance. If a paragraph is on the wrong voice, reposition the speaker label in the script. Do not export a show-notes file that still says Speaker 1.

5. 05

   Fix names and proper nouns in the script — Correct does not change the audio

   Guest spellings, product names, and town names fail first. Official Correct your transcript help: double-click the word (or the adjacent word), click Correct in the hover toolbar or press C, type the spelling, then Correct this instance or Correct All. Correct mode — Option+C on Mac, Alt+C on Windows — lets you click and type without opening the toolbar each time. Hold Z and click to cycle capitalization. Hold X and click to cycle punctuation. This is a text fix. Deleting the word would cut the media. If you want a usable transcript, Correct. If you want the ramble gone, that is a different page.

6. 06

   If the same names fail again, add a glossary — or re-transcribe the file

   Official Transcription glossary help: add proper names, brand terms, and jargon before you transcribe so the next pass recognizes them. Open the top-left menu → Tools → Transcription Glossary. One term per line. Official cap is 30 terms per Drive, English only, shared by every project on that Drive. Descript can also add a word you have corrected at least three times; that toggle lives on the same glossary window. If the whole script is the wrong language, do not Correct your way through it. Official path: file options → Re-transcribe → pick the language → Done. One spoken language per file. Mixed-language episodes are not a published workflow.

7. 07

   Export the words as text — or stay in the script and use it as the edit

   Official export help: click Export, open the Transcript tab. Published formats are .html, .md, .docx, .txt, and .rtf. Include composition name, markers, ignored text, speaker labels, or speakers in every paragraph when you need those. Timecode settings cover offset, interval, paragraph breaks, speaker labels, and markers. Wordless media and scene boundaries do not appear in the text file. That is the show-notes / blog / guest-review handoff. If the remaining job is cutting the episode in that same script, stay in the composition — the media follows the text. That radio edit is How to edit a podcast in Descript, an ordinary site path. This page already hopped to Descript.

Export text vs stay in the script vs import a transcript you already have

“Transcribe a podcast in Descript” is a search. The decision is what you walk away with. The columns match Export → Transcript, edit-like-a-doc, and Import / Replace transcript as of August 2026. We are not inventing a third product named Podcast Transcriber — that is this same editor.

Import the episode — then wait for the script

Open a project for this episode. Official getting-started copy gives you two starts: record in Descript, or import an existing file. Import paths they publish include upload from your computer, import from Zoom, and upload from a phone. Rooms records guests remotely and keeps each participant on a separate track. A mixed booth file is enough for a transcript. Separate ISOs are useful later if you will radio-edit; they are not required to get the words.

Put the file in the script. Official add-a-file help: drag it from the Project panel into the Script editor, or select it and click Insert into script. Descript then transcribes the spoken audio and aligns each word to the media. That alignment is why deleting a sentence later cuts the take. Official automatic-transcription help also lists Transcribe file on the file options menu, and automatic transcription when you import into a sequence. We describe those actions. We are not inventing a toolbar icon named Transcribe Podcast.

A file that is only a layer is not a transcript. Official copy says so. If the script is empty after import, you added media to the canvas and not to the document. Insert it. Then wait. Processing finishes, the file reaches the cloud, then the words appear. Time depends on length, audio quality, and the connection. Official help: any file longer than 15 hours fails automatic transcription. Split a multi-hour dump first. Music and lyrics are not a speech transcript — do not expect a bed or a needle-drop to come back as text.

Language is a setting, not a guess you make after the fact. Official App Settings: Default transcription language is Auto. Always ask before transcribing prompts before the pass starts; if you dismiss that picker, the file stays untranscribed. Official supported-languages help (August 2026) lists 26 Latin-alphabet languages; Chinese, Japanese, and Russian are not yet supported. English (UK) and Welsh (Beta) must be selected by hand. Confirm the live table. One spoken language per file. A bilingual interview is not a published workflow.

Media minutes start when the file is in the project. Official review-era metering we already verified on the Descript review: every file you upload or transcribe counts. A one-hour interview plus a guest ISO is two files. Do not re-import an hour you already transcribed “just to try a different model” unless you meant to spend the minutes. Change the transcription model in App Settings before the next import if you need to.

Speakers and names — the difference between a script and a usable transcript

Automatic speaker detection is on by default. Official Detect and label speakers help: Descript identifies the voices, then you name them once. When detection finishes, click Identify speakers. Add a name or pick an existing one. Close. Labels follow the file into the script. If detection was off, Detect speakers lives on the file options menu in the Project panel. We are not inventing a “Name my guest” wizard beyond that.

After the labels exist, official Speakers help is the daily surface. Type @, or select text and Change speaker, then create a name or pick one you already used. Click any instance to rename the label everywhere. If a paragraph landed on the wrong voice, hover the speaker, use the reposition control, and drop the label where that person actually starts. Show speakers on a text export is a later checkbox — it only helps if Sarah is Sarah.

Names in the body are a Correct pass, not a rewrite. Official Correct your transcript help: double-click the word, click Correct or press C, type the spelling, then this instance or Correct All. Correct mode — Option+C / Alt+C — is the published way to walk a page of proper nouns without opening the toolbar each time. Hold Z and click for capitalization. Hold X and click for punctuation. After a correction, Descript re-checks alignment; a dotted underline is the published “still aligning” state on the captions how-to, and the same script is the source.

Correct does not change the audio. That is the point of a transcript fix. Delete the word if you meant to cut the take. Ignore or Remove from transcript if you want the media to stay and the words to hide — official edit-like-a-doc help names those actions. For show notes, you want the words right and the take intact. Correct.

Glossary, re-transcribe, then export the file that left the app

If the same guest comes back next week, stop paying the Correct tax. Official Transcription glossary: top-left menu → Tools → Transcription Glossary. One name, brand, or jargon term per line. Official cap is 30 terms per Drive. English only. Shared by every project on that Drive — add it once. Business and Enterprise get extra team glossary controls in Brand Studio; we describe that as a plan feature, not a checkout we invent. Descript can add a word you have corrected at least three times; the toggle is on the same window. Add the names before the next import.

If the whole script is the wrong language, Re-transcribe. Official path: Project panel → Files → file options → Re-transcribe → pick the language → Done. Do not Correct 40 minutes of French that came back as English. Change the default language, or turn on Always ask, so the next file does not repeat it. Official help: one language per file.

If you already paid a human for a transcript, official automatic-transcription help points at Import transcript and Replace transcript to swap Descript’s automatic pass for that text. Those actions are English-only on the published language list, with filler detection, the glossary, and several other speech tools. We describe the action. We do not invent a “paste Otter into the timeline” button.

Then leave with the words, or stay. Official transcript export: Export → Transcript tab. Formats are .html, .md, .docx, .txt, and .rtf. Turn on speaker labels when the file is an interview. Turn on speakers in every paragraph if your CMS wants a name on each block. Set timecode if a producer will jump to a quote — offset, interval, paragraph breaks, speaker labels, markers. Official caveat: wordless media and scene boundaries do not appear in the text file. Listen to nothing here. Read the export. If Speaker 2 is still in it, go back.

That text file is not a caption sidecar. SRT and VTT live on Export → Subtitles, which is How to add captions in Descript. It is also not the RSS enclosure. If you stay in the script and delete a ramble, the media follows — that radio edit is How to edit a podcast in Descript. Both are ordinary site paths. This page already hops to Descript.

Free-plan media minutes are 60 a month on the August 2026 grid. That is a short episode, not a season. Sit on a paid plan before the weekly hour is the file you transcribe. Confirm checkout on descript.com/pricing. We do not print a commission rate.

When the next job is the edit, the cleanup, silence, or captions

Stay here when you need the words. Hand the composition to the radio edit when a story still has to leave: How to edit a podcast in Descript. Hand it to cleanup when the room is the problem: How to clean up podcast audio in Descript. Hand it to the silence pass when the remaining problem is time between words: How to remove silence in Descript. Hand it to captions when the picture needs type: How to add captions in Descript. None of those are Try buttons on this page. The only hop here is Descript.

Finding 5–15 verticals in a finished episode is a clipper job — start at Descript vs Klap if you have not picked that stack yet. A licensed intro, outro, or bed is How to add AI music to a podcast. Klap and Mubert stay foils here, not hops.

We do not print an affiliate commission rate, and we do not invent a “this will earn you $X” line. The product question is whether the script you export still says the guest’s name after you Corrected it — and whether you were honest about the file that never made it into the script.