ZylcutHow auto captions work in Zylcut
Add a captions layer and Zylcut transcribes the audio beneath it locally, with a start and end time for each word. Captions then follow the speech word by word. When your agent cuts out the pauses, it times the captions to each remaining piece, so they stay in sync with what is left.
- Import your video into a Zylcut project.
- Add captions with the captions button in the Layers panel, or ask your agent: "Caption this in the spotlight style."
- Pick a style and adjust the font, size, colors, position and background.
- Export with the captions rendered in, using the YouTube, Shorts, Reels or TikTok preset.
Seven caption styles
- classic: One word at a time, centred. Clean and readable.
- spotlight: A few words, with the word being spoken in an accent color. The look of most Shorts and Reels.
- punch: Bold uppercase lines; the line being spoken grows and changes color.
- cascade: Words build up line by line, left-aligned, as they are spoken.
- whisper: Calm subtitle-style lines at the bottom of the frame.
- paper: Two light lines, with the line being spoken turning bold.
- stark: Heavy uppercase words that invert the picture beneath them.
Every style can be restyled: change the font, weight, case, colors, outline, shadow and position, add a background box, or animate words as they appear.
Made for short-form and long-form
- Shorts, Reels and TikTok: big, bold, word-by-word captions on a 9:16 vertical frame.
- YouTube and tutorials: calm subtitle lines at the bottom of a 16:9 frame.
- Podcasts and interviews: transcripts with speaker labels, so each voice can be styled.
Captions plus a clean cut
The same word timings that drive the captions also let Zylcut remove silences and filler words without cutting words in half. Ask your agent for both at once: "Cut the pauses and caption every word." Learn more about editing with AI in Zylcut.
Questions
How do I add captions to a video automatically for free?
Install Zylcut (free, no watermark), import your video, and add captions from the editor or ask your agent to caption it. Zylcut transcribes the audio on your computer and syncs each word, then you pick a style and export.
Are the captions burned into the video?
Yes. Captions are part of the picture and are rendered into the exported video, which is what Shorts, Reels and TikTok need. You can also use an existing .srt or .vtt file as the source.
How accurate is the timing?
In English, every word gets a measured start and end time, so captions appear exactly as each word is said. Other languages use estimated word times, which suit captions grouped by phrase.
Which languages are supported?
English, plus 99 more through a local Whisper model. Speaker labels for interviews and podcasts are available too.
Can I fix a wrong word?
Yes. The transcript is a plain file in your project, so you or your agent can correct a word, and the captions update.
Is my audio uploaded for transcription?
No. Transcription runs entirely on your computer. The model downloads once, and after that it works offline.