SubGrab

How to Save a Video Transcript as Word, PDF or Markdown

Download a video's transcript as a Word document, PDF, Markdown notes, a spreadsheet or JSON, with or without timestamps, and pick the format for the job.

Short answer

To save a video's transcript as Word, PDF or Markdown, open the transcript on SubGrab, click Download and pick the format under Documents. Word and PDF are laid out in paragraphs with speakers and times, Markdown suits note apps, and an Include timestamps switch keeps or drops the times. SRT, WebVTT, CSV and JSON are in the same menu.

A transcript has to fit the next job. An editor wants a document to mark up, a report wants a PDF, notes belong in Markdown, a video editor wants subtitles and a spreadsheet wants rows. SubGrab's Download menu makes each of those from the same transcript, so the words are identical in every file.

How to download a transcript as Word, PDF or Markdown

1. Get the transcript: paste the video's link on the home page. Captions are used when the video has them, AI transcription when it doesn't

2. On the transcript page, open Download

3. Under Documents, set Include timestamps the way you want it, then choose Word (.docx), PDF, Markdown (.md) or Plain Text (.txt)

Your browser remembers the timestamps choice for next time. The same menu has SubRip (.srt) and WebVTT (.vtt) under Subtitles, and Spreadsheet (.csv) and JSON (.json) under Data. To paste the text somewhere instead, use Copy.

Which format for which job

FormatBest forWhat you get
Word (.docx)Editing, comments, sharing with a teamThe title, then the source link, length and language, then paragraphs with speaker and start time
PDFReading, printing, attaching to a reportThe same document, printed by your browser
Markdown (.md)Notion, Obsidian, GitHub, AI toolsA heading, then one paragraph per block of speech
Plain Text (.txt)Anything that accepts textOne timed line per caption, or paragraphs without times
SubRip (.srt), WebVTT (.vtt)Video editors, players, YouTubeTimed subtitle cues
Spreadsheet (.csv)Excel, Google Sheets, NumbersOne row per caption: start, end, speaker, text
JSON (.json)Scripts and appsTitle, link, length, language, speakers and every timed segment

Why the Word and PDF files read like documents

Captions arrive as short timed lines of a few words each, and pasted into a document they read as a stack of fragments. The Word and PDF exports join them into paragraphs instead. A paragraph closes at the end of a sentence once it reaches about 90 words and never runs past 150, and a new one starts whenever the speaker changes or the audio pauses for more than four seconds. With timestamps on, each paragraph opens with its start time, and with speaker labels it names who is talking.

Speaker names carry into every file

Rename Speaker 1 and Speaker 2 once on the transcript page and every download uses the names: Word, PDF, Markdown, text, both subtitle formats, the spreadsheet and the JSON. The steps are in how to replace Speaker 1 and Speaker 2 with real names.

Any language, any script

More than a third of SubGrab's AI transcripts are not in English, so the documents assume nothing about the alphabet. Chinese and Japanese lines are joined without stray spaces, Korean keeps its spaces, and every paragraph in Arabic, Hebrew or Urdu is set right to left on its own. File names keep the letters of the video's title in any script. The spreadsheet starts with a byte-order mark, which is what makes Excel read it as UTF-8 instead of garbling non-Latin text.

A whole list in one download

Running a list of up to 50 links ends with one ZIP: a file per finished transcript in the format you choose, an all-transcripts.md with every transcript under its own heading (in parts of at most 400,000 words, so each fits an AI tool's file limit) and a manifest.csv listing each video's title, link, platform and length. PDF is the one format a list leaves out, because a PDF is made by your browser's print dialog, one document at a time.

The formats, from their sources

  • WebVTT is the W3C's format for timed text on the web, the one HTML video players read: the WebVTT specification
  • CSV is described in RFC 4180
  • JSON is defined by RFC 8259
  • Markdown began as John Gruber's syntax, and many note apps read it today
  • Word (.docx) files open in Microsoft Word, Google Docs and Apple Pages

Get a transcript and download it your way — 2 free credits on signup →

Frequently asked

Does the PDF keep Korean, Chinese or Arabic text intact?

Yes. The documents are built for every script: Chinese and Japanese are joined without stray spaces, Korean keeps its spaces, and each paragraph in Arabic, Hebrew or Urdu is set right to left on its own. The PDF is printed by your browser with your system's fonts, and file names keep the title's own letters.

Why does the PDF open a print window instead of downloading a file?

Because your browser makes the PDF. SubGrab opens a print-ready page in a new tab and the browser's print dialog offers Save as PDF. If nothing opens, the browser blocked the pop-up: allow pop-ups for subgrab.com, or download the Word file instead, which opens in Word, Google Docs and Pages.

Can I download every transcript from a list in one go?

Yes. A finished list downloads as one ZIP with a file per transcript in the format you choose, an all-transcripts.md holding every transcript under its own heading, and a manifest.csv listing each video. PDF is the one format a list cannot include, because each PDF is made by your browser's print dialog.

Written by the SubGrab team. About SubGrab says who we are and where your audio goes; questions and corrections go to help@subgrab.com.