Story and reel covers
One word behind the person, big enough to fill the frame. The subject stepping in front of the type is what makes a cover stop a thumb.
Put a word behind the person, pet or product in your photo. The subject is cut out on your device and laid back over the type, so the effect looks printed on the wall behind them — no upload, no account.
Drop a photo to put text behind its subject
Works best with one clear subject — a person, a pet, a product. JPG, PNG, WebP or HEIC.
Text behind the subject is the poster effect of the last few years — a big word across the middle of a photo, disappearing behind a shoulder and coming back out the other side. Done by hand it is a two-minute job in Photoshop: duplicate the layer, mask the subject, put the type between the two. Done badly it is obvious immediately, because the edge of the mask never quite matches the edge of the person.
This tool does the same three layers, with the mask produced by the same segmentation model the background remover uses — running in your browser, on your photo, with nothing uploaded. Because the subject on top is cut from the identical file at the identical size, the silhouette lines up exactly, which is the whole reason the effect reads as real rather than as a sticker.
What's left is the typography, which is where the result is actually won or lost: a font that suits the picture, a size that commits, tracking wide enough to breathe, and the word placed so the subject interrupts it somewhere interesting — not dead centre.
One word behind the person, big enough to fill the frame. The subject stepping in front of the type is what makes a cover stop a thumb.
A name, a date or a city behind the headline act, printed-on-the-wall style, without opening a design app.
The product name behind the product. It reads as packaging design rather than as a caption pasted on afterwards.
A YouTube thumbnail or a LinkedIn banner where your name sits behind you instead of floating in a corner.
The subject is cut out automatically. The first run downloads the model once; after that it is instant.
Pick a font, set the size and tracking, then drag the text to where the subject should interrupt it.
Export a PNG or JPG. Free saves up to 1200px on the longest edge; Pro saves at your photo's own resolution.
No. The cutout model is downloaded to your browser and the photo is processed there. You can watch the network panel while it works — nothing leaves the machine.
It downloads the model, roughly 5 MB, once. Your browser caches it, so every photo after that starts immediately, including offline.
Yes — the font list includes Noto Sans SC, Noto Serif SC and several display faces, and the export waits for the font to load before drawing so the file matches the preview.
Add as many text layers as you like; each has its own font, size, rotation and position, and all of them sit behind the subject.
The tool says so and still works as a text-over-photo editor. It never quietly uploads the photo instead.
Free exports are capped at 1200px on the longest edge, which covers stories, feeds and thumbnails. Pro exports at the source resolution for print and wallpaper.