Combine photos

ai photo combinerSmart layouts that keep every face.

The AI finds faces in your photos and suggests layouts that never crop one.

or drop them here · paste with Ctrl+V

Free · No account · Your photos stay in this tab

Drop your photos here

Two or twenty, any sizes. Paste works too.

The order here will be the order on the board. Drag to change it.

0 picturesLaid out in this tab

An AI photo combiner that works on your device: it finds the faces in your photos and suggests layouts that keep every one of them in view. It never draws or invents anything.

Add photos, then Smart layouts: one press finds faces (a 224 KB model, after you agree) and shows the three layouts that cut the fewest.

The board is empty: add pictures or open the sample

AI that arranges, and does nothing else

Many AI photo combiners send your pictures to a server and generate a new image from them. This one is an AI photo combiner of a different kind: a small face-finding model comes to your browser, finds the faces, and the board uses them to choose a layout that crops nothing important. The pictures you get back are your own, whole or neatly cropped — not a model's idea of them.

Getting a smart layout

  1. Add your photosTwo to forty-eight, any shapes.
  2. Pick an output shapeAny, square, portrait 4:5, story 9:16 or wide 16:9.
  3. Let it look for facesOne press, after you agree to the 224 KB download.
  4. Apply a suggestionThen adjust anything — every setting stays live.

Where it stops

The model finds faces, not bodies or objects, so a crop can still cut a hand or a pet. It never generates pixels: no backgrounds are extended, no faces merged, no 'AI fusion'. Suggestions are scored, not guaranteed — look at each before you export.

Questions people ask

What does the AI actually do?

It finds faces in your photos with a small face-finding model (YuNet) that runs in your browser. The board then tries every layout it offers — rows, stacks, grids with different cell shapes, packed rows — scores each on faces cut, picture area cropped, enlargement and output shape, and shows you the best three, drawn from your own photos.

Does it generate or change my photos?

No. It does not draw, extend, swap faces or invent anything. Every suggestion is an ordinary layout of your photos, which you can change like any other.

How big is the download?

The face finder is 229,738 bytes (about 224 KB), downloaded once from our model file server after you agree, plus the runtime that runs it (about 13.4 MB, from this site). Both are kept in your browser for next time; Clear downloaded models removes them.

Are my photos uploaded to an AI service?

No. The model comes to your device; your photos never go anywhere. Faces are found in this tab.

Can I turn the AI off?

It is never on until you ask. Suggest without looking for faces scores layouts on cropping and shape alone, and every layout control works without any model.

What if it finds no faces?

Then the suggestions weigh cropping, enlargement and shape only, and say so. Landscapes, products and documents are fine.

Nearby

Photo gridSet the rows and columns yourself.Combine multiple photosUp to 48 photos packed without cropping.Silhouette double exposureOn-device AI that finds the person.