SRT to Word Guides

Descript → Word

Descript Export to Word

Descript can hand you a transcript document or a subtitle file. Open the document if you have it. Drop the SRT or VTT here when the client asked for a Word script built from the caption cues.

Do this

All guides
  • Show timestamps On to check the composition, off for a reading script
  • Merge same speaker On after speaker labels exist
  1. See which file Descript actually gave you A .docx or .txt transcript is already a document. A .srt or .vtt is cues. This converter reads the cues. It does not open a Descript project.
  2. Expect the edit you already made Filler-word removal, studio sound, and Overdub live in Descript before export. The SRT contains the words that remained. This site does not restore words you deleted, and it does not remove new ones.
  3. Check the speaker label at the start of the cue Labels that look like [Alex] or Alex: become bold names. A subtitle export with no label merges nearby lines into one voice. Add the label in the SRT if the preview never changes speaker.
  4. Drop the SRT with timestamps on The preview should be paragraphs, not cue numbers. If you still see --> lines, the file did not parse — do not download that view.
  5. Download the .docx and leave the composition alone Word does not push edits back into Descript. Change lines that must play in the video inside Descript. Change the client script in the .docx.

Drop the subtitle export, not a screen recording. Show timestamps on if you will check a sentence against the composition. Merge same speaker on joins labeled turns. Download descript-script.docx and keep the SRT.

Two exports, and only one belongs here

Descript is an editor. It can give you a transcript document — often already .docx, text, or Markdown — and it can give you subtitles as .srt or .vtt for a player. People land on “Descript to Word” after they exported the captions for YouTube or a rough cut and then needed a script the producer can comment on. If the transcript document is already on disk, opening it is the whole job. Dropping that .docx into this converter will not parse: there are no cue arrows in a Word file.

The file that does belong here is the subtitle export. Open it. You should see cue numbers or a WEBVTT header and --> lines. That is timed text chosen for the frame, which is a worse reading copy than Descript’s own transcript and a better reading copy than the raw SRT. The converter removes the cue furniture and writes paragraphs.

What already happened inside Descript

Filler-word removal, studio sound, and Overdub are edits to the composition. They are finished before export. The SRT is a photograph of the words that remained. This site cannot put “um” back, cannot hear which take you kept, and cannot tell an Overdubbed sentence from a sentence someone actually said. If the client must know which lines were replaced, mark that in Descript or in the Word file yourself. Do not expect a badge in the .docx.

Speaker names follow the cue text. A label at the start — [Alex] or Alex: — becomes a bold name, and Merge same speaker joins that person’s nearby cues. A subtitle pass that was burned for a single narrator often has no labels. Merge then produces one voice, which is what you want for a monologue and the wrong document for an interview. Add labels before you convert, or follow interview SRT to Word if the file is really a two-hander.

Check the composition, then make the reading copy

Drop the SRT with timestamps on. The ranges in the preview are a short form of the cue clocks. Spot-check one line against the Descript composition. If the words match and the range is in the right minute, download descript-script.docx. For a client who should read prose, use the timestamps-off link and download again. Keep the SRT if those captions are also going to a player. Descript will not import your Word comments back onto the script track.

Otter and Teams exports look similar and are different files. Use Otter transcript to Word when the source is Otter, including when the paste is a summary glued to the transcript. Do not reconvert a Descript .docx “to be safe.” You would be starting from the wrong object.

Related workflows

Download the .docx

Drop the subtitle export, not a screen recording. Show timestamps on if you will check a sentence against the composition. Merge same speaker on joins labeled turns. Download descript-script.docx and keep the SRT.

Descript export questions

Which Descript file is already Word, and which one still needs a convert.

Descript already exported a .docx. Should I run it through here?

No. Open that file. This converter reads SRT, VTT, ASS, and speaker-plus-timecode text. It does not open a Word document or a Descript project.

Will Overdub or removed filler words be undone?

No. The subtitle export contains the words Descript wrote into the cues after those edits. This site does not talk to your Descript account and does not restore audio you changed.

Why did two speakers become one paragraph?

The cue text had no [Name] or Name: label, so nearby lines count as the same voice. Add the label in the SRT, spelled the same way each time, and convert again. This tool does not listen to the composition.

Can I push Word edits back into Descript?

No. Change lines that must play in the video inside Descript, then export again if the script should match. The .docx is a document, not a project file.