Edit SRT and VTT Subtitles Online
Import SRT or VTT captions on this page, edit cue text and start or end times in a table, validate the timing, and download a saved subtitle version. This subtitle editor does not run speech recognition and does not use transcription credits.
Limits: 2 MiB UTF-8 and 10,000 cues. Editing does not run speech recognition and does not use transcription credits.
Exports are built in this browser from the caption text on this page. Sign in only if you want the saved version in My records.
Editing parses caption text and runs in the browser. Saving a signed-in task stores the original captions and the edited version on the account — that save is not an offline-only promise. Editing does not transcribe audio and does not use credits. See our privacy policy and plans and usage limits.
How to edit SRT and VTT subtitles
Import a caption file, fix what the player would show, and export the corrected track. Nothing here sends audio through speech recognition.
- 1
Import SRT or VTT
Choose a .srt or .vtt file, paste the caption text, or load the public sample. The page reads UTF-8 SubRip and WebVTT cues and fills the editor table with start time, end time, and caption text for every cue.
- 2
Edit the cue table
Rewrite caption text, type new start and end times, add a cue to the end, delete a row, undo a change, or find-and-replace a word across every cue. Overlapping rows are flagged but never deleted.
- 3
Validate, download, or save
Download SRT or VTT whenever the timing is valid and every cue has text. Sign in and save a version to reopen the same cue table later from My records. Saving stores the original captions and the edited version.
A labeled before and after edit
This sample is an original caption set we made for this page, not a customer file and not an accuracy score.
SRT input with an overlap
1
00:00:00,000 --> 00:00:02,400
Welcome back to the product demo.
2
00:00:02,200 --> 00:00:04,600
This cue overlaps cue 1 on purpose.
Both rows stay visible in the editor.
3
00:00:05,000 --> 00:00:08,200
Take hallway 3 → room 12.
The door code is 4815.Cue 2 starts at 00:00:02,200 while cue 1 still runs until 00:00:02,400. The editor keeps both rows and shows an overlap warning instead of guessing.
Download sample SRTEdited SRT after fixing the timing
1
00:00:00,000 --> 00:00:02,400
Welcome back to the product demo.
2
00:00:02,500 --> 00:00:04,600
This cue now starts after cue 1 ends.
Both rows stay visible in the editor.
3
00:00:05,000 --> 00:00:08,200
Take hallway 3 → room 12.
The door code is 4815.Cue 2 now starts at 00:00:02,500, after cue 1 ends. The arrow and room number in cue 3 are ordinary caption characters, not timing metadata.
Edit text and timecodes
Every cue is one row with three editable fields: start, end, and the words that appear on screen.
An SRT or VTT file is a sequence of timed caption cues. The editor turns that sequence into a table: each row has a start time, an end time, and the caption text. Change any of the three and the row updates immediately; exports later use exactly what the table holds now.
Add a cue to the end of the table when a line is missing, or remove a row that should not appear. Undo steps back through your recent changes, and find-and-replace updates a repeated word across every cue in one pass instead of clicking through the whole file.
Because the table is text, you can edit SRT and VTT without the original video and without sound. Caption text is stored as plain text; the editor does not re-read timing from a waveform, because there is no waveform here.
Validate caption timing
Downloadable subtitles need readable times and caption text. The editor refuses invalid timing and warns about overlaps instead of hiding them.
Each cue must have integer start and end times in milliseconds, an end time later than the start, and at least one caption character. If any cue fails those checks, the Download SRT and Download VTT buttons report the exact row and stay disabled until it is fixed. The save endpoint enforces the same rules server-side, so the browser is not the only boundary.
Overlapping cues are a separate, softer case: they are legal SRT and VTT, and many caption files contain them from repeated takes or overlapping speakers. The page shows an overlap warning and keeps both rows. It never deletes or re-times a cue on its own.
The simple number format takes 00:00:01,000 or 00:00:01.000. Two-digit hours and minutes plus three millisecond digits are accepted, and the exported SRT always uses commas while VTT uses dots.
Save and export versions
Download SRT, download VTT, or keep a saved version in My records and reopen it after a refresh.
Exports are built in this browser from the cues in the table. SRT keeps the numbered SubRip shape most editors and platforms import; VTT adds the WEBVTT header used by HTML5 players and streaming platforms. You do not have to pick one — download both from the same edited table.
Signed-in users can save a version. The record keeps the original caption file and the edited cue table separately, so you can see what changed. Reopen the task from My records or the task link and the page restores your saved cue table without re-typing anything.
Editing a caption file never starts a transcription, does not use transcription minutes, and does not change the audio or video it came from. If you need new captions from a recording instead, capture them first with the SRT or VTT subtitle generator, then come back here to polish the text and timing.
WhisperWeb vs desktop editors vs cloud caption studios
Choose the workflow that matches the task. Public product pages checked 19 September 2026; individual plan details change.
| Decision | WhisperWeb | Desktop subtitle editor | Cloud caption studio |
|---|---|---|---|
| What you start with | WhisperWebAvailable: An existing SRT or VTT fileImport a caption file you already exported from a video editor, platform, or generator. No video upload is required. | Desktop subtitle editorAvailable: Local caption filesInstall an editor on your machine, then export SRT or VTT from it. Timing and styling tools are local. | Cloud caption studioLimited: Upload to a hosted editorMany studios pair caption editing with paid transcription, so importing a finished file can sit behind an upsell. |
| Cost and speech recognition | WhisperWebAvailable: No transcription creditsThis page parses and edits caption text. It never runs a speech model and never bills minutes. | Desktop subtitle editorAvailable: No per-minute chargeThe installed app edits text locally. You pay for the software and your own machine time. | Cloud caption studioNot available: Often a billed ASR pathCloud studios commonly charge when they generate captions from audio, even for a short clip. |
| Versions and records | WhisperWebAvailable: Saved versions reopen hereSave the original file and the edited cue table to My records. Refresh or reopen later without re-typing. | Desktop subtitle editorLimited: Project files on diskUndo and project files live on your computer. There is no saved WhisperWeb task URL to reopen. | Cloud caption studioLimited: Files on a vendor hostRetention windows, downloads, and privacy terms belong to that host, not to a page you control. |
WhisperWeb details match the editor on this page. Desktop and cloud columns describe common approaches seen on public caption editor pages on 19 September 2026, not a paid test of every vendor. See our plans and usage limits for billed speech tools, which this page does not use.
Use WhisperWeb when you already hold SRT or VTT and need a browser table with undo, search-and-replace, validation, and both formats from the same edits. Use a desktop editor when you need waveform retiming, styling, or advanced positioning that this page does not offer. Use a cloud caption studio when the real job is creating captions from a recording, then come back here to edit the result.
From a caption track to a publishable version
The editor keeps the time-coded structure so the result still imports into players, editors, and platforms.
Correct the words and the clocks in one pass
Captions are shown, then forgotten — mistakes like a mis-split line, a wrong start time, or a repeated sentence read loudly. A subtitle editor exists so those fixes happen in one pass over the cue table instead of re-exporting from a video editor. Download the corrected track and import it back into your platform.
Generated use-case illustration, not an actual customer result. The runnable sample above is separate and downloadable.
Limits, privacy, and cost
Accepted input is UTF-8 SubRip (.srt) or WebVTT (.vtt), uploaded or pasted. The limit is 2 MB and 10,000 cues. The browser and the save endpoint both enforce those checks, plus integer millisecond times, end-after-start, and non-empty caption text. AUDIO, video, ASS, and SBV files are rejected on this page.
Editing does not upload media to a speech model and does not deduct transcription minutes. Guests import, edit, and download in this browser. A signed-in save stores the original captions and the edited cue table under the owning account so a refresh can reopen the task. A File object cannot be reconstructed after the tab is gone; if the original upload is missing, the page asks you to choose the file again.
Private files stay owner-only through the existing task read path. Original captions and edited versions are not placed in a public index, and the site does not log caption bodies. Do not upload material you are not allowed to process. Editing captions does not re-sync a video; download the track and import it into your editor or platform.
Frequently asked questions
Can I edit without the original video?
Yes. The page reads only the caption text. SRT and VTT contain start and end times plus the words, so every edit happens in the cue table without the audio or video file. You never upload media here.
Can I undo changes?
Yes. Undo steps back through your recent edits, including text, timing, added, and deleted rows. It works within the current editing session. A saved version stores the full cue table, so reopening a saved task restores exactly what you saved.
Does this automatically sync to speech?
No. The editor never listens to audio and never changes timing on its own. It re-times nothing automatically; overlap warnings tell you a cue starts early so a human can fix it. To create captions from a recording, use the SRT or VTT subtitle generator first.
Can I add or remove subtitle lines?
Yes. Add cue appends a blank row to the end of the table, and the trash button removes a row. Empty rows with no caption text block export until you fill them in.
Is an overlap a download error?
No. Overlapping cues are legal SRT and VTT, so the page keeps both rows and shows a warning. Export is blocked only for invalid timing, such as an end time at or before the start, or an empty caption.
Who can reopen a saved subtitle version?
Only the owning account. After you save, the task link on this page and the row in My records both restore the original captions and the edited cue table. Another account that opens the same URL does not receive the file. Guests can download without saving.