mimicThree ways to lay it out
| Layout | Best for |
|---|---|
| Paragraphs | Blog posts, show notes, newsletters. A new paragraph starts after a pause you set (2 s by default). |
| One per caption | Checking a script against what was said, with optional [m:ss] timestamps. |
| One block | Pasting into an AI tool, a word counter or a caption. |
What gets cleaned
Caption numbers and timecodes are removed. Line breaks inside a caption become spaces so sentences read straight through. Tags like <i>, <b>, <c.yellow> and {\an8} positioning codes are stripped.
Turn on Remove speaker labels to drop sound cues and names at the start of a line: [MUSIC], (laughs), JOHN:.
Worked example
Three captions: “I tried every serum on my shelf” (0:00–0:02.4), “and this is the only one / I actually finished.” (0:02.4–0:05.1), “Here's why it works.” (0:05.3).
Paragraphs, 2 s pause: one paragraph, since the longest gap is 0.2 s.
Result: I tried every serum on my shelf and this is the only one I actually finished. Here's why it works.
Questions
How do I convert an SRT file to text?
Pick or paste the file, choose a layout, then copy or download the .txt. Numbers and timecodes are removed for you.
Can I keep timestamps?
Yes. Turn on Timestamps to start each caption or paragraph with [m:ss], handy for show notes and YouTube descriptions.
Does it work with YouTube's auto captions?
Yes, if you download them as .srt or .vtt from YouTube Studio. Auto captions have no punctuation, so the paragraphs mode (which splits at pauses) reads best.
Is my transcript uploaded?
No. It's made in your browser.