SRT to TXT – Extract a Transcript
Strip the sequence numbers, timestamps and formatting tags from a subtitle file and keep only the words. The result is a clean transcript you can paste into a document, feed to a translation tool or publish as show notes. Works with .srt, .vtt and .sbv input.
Accepted: .srt, .vtt, .sbv · You can add several files at once.
Your files never leave your device.
How to use this tool
- Drop your subtitle file(s) on the box above.
- Open the preview to confirm the text looks right.
- Download the .txt transcript.
What you get
Each cue becomes one line of text. Line breaks inside a cue — which exist only to fit the text on screen — are joined with a space, and inline tags such as <i>, <b>, <c.class> and <v Speaker> are removed. HTML entities like & are decoded. Empty cues are dropped so the transcript has no blank lines.
The output is UTF-8 plain text that opens in any editor, word processor or note-taking app.
When is a plain-text transcript useful?
Transcripts are the fastest way to turn a video into searchable, quotable content. Common uses include writing a blog post or show notes from a recording, creating study material from a lecture, running the text through a translation service or large language model, indexing a video library for search, or proofreading dialogue without the distraction of timecodes. Accessibility guidelines also recommend offering a transcript alongside captions.
Keeping the structure
Because cues are usually sentence fragments, the transcript keeps one cue per line rather than guessing at paragraphs. This makes it easy to re-flow in a word processor (replace line breaks with spaces) while retaining the ability to map a line back to its cue. If you later need the timing again, keep the original subtitle file; the conversion is one-way. For the reverse direction, convert text into subtitles using a dedicated captioning tool.
Frequently asked questions
- Can I keep the timestamps in the transcript?
- This page produces a clean text-only transcript. The core library supports a timestamped variant ([m:ss] prefix per line); if you need it, open an issue and it can be exposed as an option.
- Does it work with VTT and SBV files too?
- Yes. The format is detected from the content, so you can drop WebVTT and SBV files on this page and get the same plain-text output.
- Will speaker names be kept?
- Names written as plain text (e.g. “JOHN: Hello”) are kept. WebVTT voice tags like <v John> are structural markup and are removed, leaving only the spoken words.
- Is my subtitle file uploaded?
- No. The extraction runs in your browser and the file is never transmitted anywhere.