Auto captions timed to every word you say

What you get

The Captions tab holds everything. Turn on "Show captions" and these are yours to set:

• Word timing from the transcript. Each caption shows 1 to 3 words, set with "Words at a time".
• 8 styles and 7 templates: Word pop, Karaoke, Keywords, Keywords and emoji, Two lines, Outline and Boxed.
• "Highlight the spoken word", which lights up each word as it's said.
• Keyword captions, which put numbers, names and strong words in your accent color.
• A new line after a comma, so captions break where your sentence breaks.
• Position at the top, middle or bottom, or drag the caption to the exact spot.
• "Download captions (.srt)" for a separate subtitle file.

How to add captions

1. 01Import your recording and let the analysis finish. Captions are built from the transcript, so there's no separate captioning step to wait for.
2. 02Open the Captions tab and turn on "Show captions".
3. 03Pick a template, then a style. Set the words at a time and the size.
4. 04Play it back. To fix a misheard word, double-click it in the transcript and type the right spelling.
5. 05Drag the caption on the preview if it covers your face or the product you're holding.
6. 06Export the MP4 with captions burned in, or download the SRT for YouTube or a course platform.

Remembered spellings

Speech models stumble on names, brands and jargon. If your bakery is called Krumb and the transcript keeps writing "crumb", fix it once. Cutroom AI remembers the spelling and uses it in your new edits, so you aren't correcting the same word every week.

This pays off most for people who say the same unusual words often, like a coach's program name or a pastor's sermon series.

Captions on clips and versions

Clips cut with the clip finder keep their captions, and the crop follows your face, so a vertical clip pulled from a horizontal talk still reads well. Each version gets its own caption treatment: plain captions on the natural cut, keyword captions on the tight cut. Change one version's captions and the others stay as they were.

What it can't do

It doesn't translate captions. Caption translation is one of the cloud text features, and those are switched off. Captions come out in the language you spoke, in any of the 16 languages it transcribes.

It has fewer caption styles than tools that sell big style libraries. If you want a hundred animated looks, this isn't that.

Captions follow the transcript, so a mumbled word can come out wrong. Read them once before you export, especially numbers and names.

When another approach is better

Uploading a long video to YouTube? Burned-in captions aren't needed there. Export the video without them and upload the SRT next to it, so viewers can switch captions on or off.

If you need captions in a second language, run the exported SRT through a translation service of your choice.