Getting the words out of a recording is only half the job. The other half is getting them into a form the next person, or the next piece of software, can actually use. A perfect transcript in the wrong format is still a problem.
PeekoType exports five, and choosing between them is simpler than it looks once you know what each is for.
The short version: TXT for text you will paste somewhere else. SRT for video editors. VTT for websites. DOCX when a human will edit it. PDF when a human will read it and nobody should change it. All five are produced offline and saved next to the original file.
TXT, the honest default
Plain text with no formatting and no timestamps. Nothing opens it badly, nothing mangles it, and it pastes cleanly into anything.
Choose it when the transcript is raw material rather than a finished document: something you will paste into a report, search through, feed into another tool, or read once and discard. It is the right answer far more often than people expect.
SRT, for video editors
SubRip files are timed subtitles: numbered blocks, each with a start and end time and a line or two of text. They are the closest thing subtitling has to a universal standard.
Choose SRT when the transcript is going into video software. Premiere, DaVinci Resolve, Final Cut and essentially every editor made in the last twenty years will import it and lay the captions onto your timeline. YouTube accepts it directly too, which is the fastest route to captions on an existing upload.
VTT, for the web
WebVTT does the same job with a slightly different syntax, and it is the format HTML5 video expects. If your video is going onto your own website inside a video tag, this is the one.
The practical rule: if the destination is a video editor, use SRT. If the destination is a web page, use VTT. If you are unsure, SRT is more widely accepted.
DOCX, when a person will edit it
A real Word document, ready to be opened, marked up and tracked. Choose it when the transcript is going to a colleague, a client, a supervisor or a court bundle, and somebody will need to comment on it or change it.
This is the format that makes interview transcripts genuinely usable in research and legal work, particularly with speaker labels turned on so it reads as a conversation. Our guide to transcribing interviews with speaker detection covers that side of it.
PDF, when it should stay put
Fixed layout, looks the same everywhere, awkward to edit. All of which is exactly what you want for a record.
Choose PDF for something being filed, archived, attached to a case, submitted, or printed. The awkwardness of editing is the feature, not a limitation.
Where the files go
Each export is saved next to the file it came from, using the same name with a different extension. Transcribe interview.mp3 as SRT and you get interview.srt in the same folder.
This matters more than it sounds when you are batch transcribing. Drop a folder of thirty recordings in, come back later, and every transcript is sitting beside its source rather than piled into a downloads folder in the order they happened to finish. If you also translate, the translated export carries a language code in its name so the two never collide.
Choosing quickly
- Pasting it somewhere: TXT
- Captions in a video editor: SRT
- Captions on a web page: VTT
- Someone will edit or comment: DOCX
- Filing, printing or sending as a record: PDF
You can transcribe the same file twice with different formats if you need both, and since it all runs locally there is no per-minute charge for doing so. That is part of the wider argument for keeping everything on your own machine.
The Transcribe tab and all its settings are documented in the user guide.