▰
▰
Subtrims.
Start creating ↗
FREE / ENGLISH SUBTITLES

How to add subtitles to YouTube Shorts for free

Subtrims is free, English-only, and runs speech-to-text locally in your browser. No credit card needed.

Generate subtitles free ↗

Why subtitle your Shorts

Shorts are vertical, fast and often watched on mute. Clear on-screen captions help viewers follow what is being said. Short captions of a few words each fit the 9:16 frame without covering the action.

Step 1: Generate the subtitles

Open Subtrims and upload your Short or its audio. Choose how many words per caption (3 is a good starting point) and generate. The English speech is transcribed on your device; nothing is uploaded to a server.

Step 2: Review and edit

Speech recognition is never perfect. Scroll the caption list and click any line to correct names, slang or misheard words. Background noise and overlapping speakers are the most common causes of errors.

Step 3: Export SRT

Download the SRT file. It contains every caption with its start and end time. If you only want the script, download TXT instead.

What the file looks like

Each caption in an SRT is a number, a start and end time (hours:minutes:seconds,milliseconds), the text, and a blank line. Here is a short illustrative example of 3-word captions where punctuation ends a caption early:

1
00:00:00,000 --> 00:00:01,200
YOU HAVE A STORY.

2
00:00:01,200 --> 00:00:02,600
MAKE EVERY WORD

3
00:00:02,600 --> 00:00:03,500
COUNT.

Step 4: Add the captions in your editor

Import the SRT into a video editor that supports it, such as Premiere Pro (as a caption track). Style the captions there, export the finished Short and upload it to YouTube. Subtrims makes subtitle files and does not render captions into your video, and SRT does not reliably carry font styling, so do the styling in your editor.

Tips for cleaner captions

Limits worth knowing

The free version handles English clips up to 5 minutes and 100 MB, with no credit card needed. There is no translation. The first run downloads the speech model, so it can take a few minutes; later runs reuse the cached model. Accuracy is best with clear speech: background noise, overlapping speakers, names and accents can cause mistakes, so skim the captions before exporting.

Generate subtitles free ↗

More guides