Add text to a video
Burn a caption or title straight into the picture, so it survives any re-upload.
Your file
Drop a file here, or click to choose
It stays on this device. Nothing is sent anywhere.
Settings
Make it readable
Text over video fails when the background changes under it. Two settings fix that: an outline of 2 to 4 pixels, which is on by default, or a background box for longer lines. White text with a black outline is readable over almost anything, which is why every broadcaster uses it.
Font size is in pixels of the source video. On a 1080p file, 48 px is a comfortable subtitle and 90 px is a title card. On a 480p file, 48 px is already very large.
What this is not
This draws one line of text over the entire clip. For timed captions that come and go, write an SRT file and use the subtitle burner instead.
Burned in, and what that means
The text becomes part of the picture. It cannot be turned off, translated, restyled or selected afterwards, and it survives everything: re-encoding, re-uploading, screen recording, and platforms that strip subtitle tracks. That permanence is the point when the text is a title, a handle or a watermark, and it is a drawback when the text is dialogue that someone might want in another language.
It also means the text is encoded at the resolution you export at. Fine text on a video that a platform later re-encodes to a lower bitrate will soften along with everything else, which is an argument for slightly larger and slightly bolder than looks necessary on your own screen.
Sizing against the source, not against your monitor
Font size is in pixels of the source video, so the same number means different things on different footage. On a 1080p file, 48 px is a comfortable subtitle, 60 to 72 px is a strong lower third, and 90 px is a title card. On a 720p file, multiply by about two thirds. On a 480p clip, 48 px is already dominating the frame.
Check the resolution first with media info rather than guessing, particularly with footage from a phone, which is often taller and larger than expected.
Where to put it
On anything vertical, keep text out of the lower fifth and the right-hand edge. That is where TikTok, Reels and Shorts draw captions, usernames, buttons and the progress bar, and text placed there is simply covered up. The safe zone in practice is the middle band of the frame.
On landscape video, the lower third is conventional and mostly free, though YouTube's own controls appear over the bottom of the frame while a viewer is scrubbing. Keeping a comfortable margin from every edge is the habit worth having: some platforms crop slightly, and text touching an edge is the first thing to be lost.
Readability, concretely
White with a black outline of 2 to 4 pixels is readable over nearly any footage, which is why broadcasters settled on it decades ago. A background box is better for more than a few words, because an outline does not stop a busy background competing with a long line. Avoid pure colours for the text itself; heavy compression treats saturated red and blue worst, and thin coloured type is where that shows first.
Common questions
Can I use my own font?
Not yet. The tool ships DejaVu Sans, which covers Latin, Greek and Cyrillic.
Can the text appear only at certain times?
Not here. Use an SRT file with the subtitle burner for timed text.
Do emoji work?
No. The bundled font has no color emoji glyphs, so they will render as blanks.