Text to SRT Converter for Written Transcripts
Text to SRT converter turns an already written transcript into numbered cues using timing rules you choose.
Up to 200,000 UTF-16 characters and 2,000 cues. Files are read in this page; no media upload or account is needed. This page loads analytics scripts, so avoid confidential source text.
One nonblank line becomes one cue. Blank lines are skipped; your line text is preserved.
Review cues
SRT follows common numbered-cue convention; see Library of Congress SRT description. Fixed times cannot reflect speech speed. Review media and edit cues before use.
Next: convert SRT and VTT, validate generated SRT or shift subtitle timestamps.
Why make SRT from written text?
Timing you can inspect
Set first start, duration and gap yourself. The preview shows the schedule instead of implying that text knows when speech occurs. Every generated time follows the same arithmetic, so a colleague can reproduce your schedule from three numbers.
One written line, one cue
Each nonblank input line becomes a numbered cue in the same order. Blank lines are skipped and Unicode text stays intact, including accents, emoji and right-to-left scripts that naive byte edits would corrupt.
Edit before export
Review cue text, starts and ends, correct a field, then copy or download only the current SRT. Nothing is written to the file until the edited cue passes the same parser that reads imported subtitles.
How text to SRT conversion works
SRT stores a numbered cue, a start and end time, and caption text. This tool starts with text you already wrote, not a recording. It assigns times from your initial start, fixed duration and chosen gap, so identical settings produce an identical schedule and the same file bytes.
For two cues with a 1,000 ms start, 2,000 ms duration and 500 ms gap, the first runs from 1 to 3 seconds and the second from 3.5 to 5.5 seconds. These are provisional editing times. Listen to media and adjust cues before relying on them, because a fixed duration cannot track how fast a speaker actually talks.
Input and timing rules
- TranscriptPlain UTF-8 text, one nonblank physical line per cue. No audio, video or transcription input. A lone carriage return is rejected rather than silently splitting a line.
- ScheduleStart plus duration sets each end; start plus duration plus gap sets the next start. Milliseconds are whole numbers, so no rounding drift builds up across a long track.
- BoundaryPositive duration and nonnegative gap are required. Manual overlap is shown for review; out-of-order starts block export.
- Cue numbersNumbers run 1, 2, 3 in file order. Most players place cues by their timecodes rather than their numbers, but editors and review tools list them by number, so a track that restarts or repeats numbers is hard to check by eye.
- Overlapping rangesA cue that starts before the previous one ends is allowed after a manual edit and flagged as an advisory. Players disagree here: some draw both lines at once, others show only the later cue, so an overlap you accept will look different in different software.
- Size limitGenerated output is refused above 200,000 characters. Split a long transcript rather than shipping a file that chokes a player or an upload form.
Common Use Cases
Draft captions from a script
Create an editable timed starting point before checking a video. Keep the script line breaks, since they become the cue boundaries.
Prepare translated lines
Turn translated dialogue into ordered cues, then revise timing against original media. Numbers give reviewers a shared reference for each line.
Prototype a subtitle track
Test cue wording and sequence before detailed timing work, then replace the provisional schedule with measured values.
How to create SRT from text
Add written lines
Paste or open plain text with one intended caption per nonblank line.
Set the schedule
Enter first start, positive duration and gap in the shown units.
Review and edit cues
Generate cues, page through up to 50 per view, edit one cue at a time and resolve errors.
Export current SRT
Copy or download validated .srt, then check it against your media.
A visible starting point
No transcription promise
The page never claims to hear speech or align words to audio, and there is no upload path that could send a recording anywhere.
Deterministic timing
You can recalculate every generated time from the displayed rule, which makes a disputed cue easy to explain during review.
Current output only
Changing source or timing invalidates previous export, so a stale download cannot be mistaken for the schedule on screen.
Before using generated captions
- •Keep lines intentional: Each nonblank line becomes a cue; split or join lines before generating.
- •Review whole track: Fixed timing will not fit every spoken phrase. Page through cues to inspect later ones rather than trusting the first page.
- •Check structure next: Use subtitle validator, then compare timing with recording. A file can parse cleanly and still be badly timed.
- •Watch overlaps after edits: A hand-edited start can cross the previous cue end. The advisory names the cue, so fix or accept it deliberately before exporting.
Turn a written transcript into SRT
Paste lines above and choose an explicit start, duration and gap. The first cue time is yours to set, so the schedule starts where your media starts.
Review editable cues and export current file, without audio transcription or automatic sync claims.
Related tools
Frequently Asked Questions
Does text to SRT converter transcribe audio?
No. It accepts text you already wrote. It has no speech recognition, audio upload or audio alignment step.
How are SRT timestamps calculated?
Each cue starts after previous cue duration and gap. End is start plus chosen duration, using milliseconds shown beside fields.
Will generated timestamps match my video?
Not automatically. They are provisional values from your settings. Edit cues and review them against media.
Can I edit a cue before downloading?
Yes. Change one cue text or timing, resolve any reported error and export only after current cue data validates.