Descript — Audio to Text

Descript

Descript turns audio or video into an editable transcript so you can cut and rearrange spoken content by editing text. Its free plan is enough for a short trial before you decide whether longer files and watermark-free export are worth paying for.

Why it fits

Audio to Text

Website

Go to descript.com
Good to know (1)

Check minute, upload and summary limits for your chosen plan and language.

Good to know
Good to knowWebsite
Free output—
No account needed—
No watermark—
Works offline—
Batch processing—
Content language—

— Unverified conditions. Official documentation reviewed. The app itself has not been hands-on tested. Matches only documented content-language support, not interface language.

Sources and review date (1)
  • www.descript.com/transcription
Founder feedback and correction

Founder feedback and correction

If you represent this product, send a source-backed correction or update to the editors. Publication follows editorial review.

Editorial selection guide

Last checked:

Who is it for?

  • Podcasters, interviewers and video makers who prefer to remove filler words and reorder sections by editing text.
  • People who need a transcript of an interview or lecture that they can correct and export as text, a Word file or subtitles.

Getting started

  • Create a new project and drag in an audio file (WAV, MP3, AAC, AIFF, M4A or FLAC) or a video file; transcription starts automatically.
  • Read the transcript while listening: check speaker labels, names, numbers and technical terms, and correct errors by highlighting the text and pressing C, which fixes the words without altering the audio.
  • Edit by deleting or moving text, and remove filler words; listen to each cut because the media changes with the text.
  • Export the transcript (plain text, rich text, markdown, HTML, Word, SRT or VTT) or the media, remembering that free exports are 720p with a watermark.

Decision points

  • Check your language: the transcription page says 30+ languages while the audio-to-text page says 23+, so test a sample in your language before relying on either number.
  • The pricing page lists a Free plan with 60 minutes of media per month, 720p watermarked export and 100 one-time AI credits; longer files, watermark-free export and more media hours need a paid plan. Figures and prices change.
  • If you only need a one-off transcript, a simpler transcription tool may be enough; Descript pays off when you also edit the audio or video by changing the text.

Limits

  • Descript advertises up to 95% accuracy, which is a vendor claim; accents, background noise and overlapping voices can raise error rates, so review names, numbers and quotes before publishing. This is an official-documentation review, not a hands-on test.
  • Media minutes and AI credits cap usage, and AI credits cover features such as Underlord, Studio Sound, Green Screen, Eye Contact and AI-generated media; plans and prices change.
  • Editing by text changes the media itself, so keep an untouched copy of the original before heavy cuts.

Frequently asked questions

Does it label different speakers?

The audio-to-text page says that when audio has more than one speaker, Descript identifies and labels each person. Check the labels wherever voices overlap and correct them by hand.

Can I use it for free?

Yes, within the Free plan limits shown on the pricing page: 60 minutes of media per month, 720p export with a watermark and 100 one-time AI credits. These figures can change, so confirm them on the current page.

Which files can I import and export?

The page lists WAV, MP3, AAC, AIFF, M4A and FLAC audio, plus video files. Transcripts can be exported as plain text, rich text, markdown, HTML, Word, SRT or VTT.

Official sources (3)
www.descript.com/transcriptionwww.descript.com/tools/audio-textwww.descript.com/pricing