
Whoever was holding the camera
One person took the shot, so one person is not in it. Add them back and the picture finally has everyone who was actually there that day.
Two photos in · One picture out · No masking by hand
Two uploads and the photo is finally complete — everyone your family wishes had been standing there, in it.
A real run of this tool, from the two photos above. Drag the divider: the woman, the young man, the girl in the yellow dress, the hedge and the low evening sun are all exactly where they were — and the man from the second upload is standing on the boards to the left, lit by the same sun, with his own shadow on the deck. Picture the group photo you have, with the person who should have been in it.


Drag to compare before and after
You hand over two pictures: the photograph that already exists, and a photo of the person who is not in it. What comes back is one picture in which they stand there with everyone else, at the right scale, in the right light, casting the right shadow. It is the request people type almost word for word — add person to photo — without opening an editor and without paying a stranger on a forum to do it by hand.
Every other way to add person to photo costs you something. A reshoot means getting everyone back into one room, which is usually the exact thing that cannot happen. A manual composite means an hour of masking and then another hour losing the argument with the shadows. Posting the job to a forum and offering twenty dollars means waiting days and taking whatever comes back.
Here it is two uploads and one sentence. The edit runs on GPT Image 2.5 Flare in the browser with free credits to start, the flow opens on image-to-image so your own photo is the starting point, and the credit estimate updates before you commit.
Creative engine
Upload the scene, upload the person, say where they go, and the credit estimate updates before you generate.

Whatever you already have. A gap in the group is not required — someone can be added at the end of a row, half behind a shoulder, or seated in a chair that is already empty.

One clear picture of their face is enough. It does not need to match the lighting or the angle of the first photo; matching those is the whole job you are handing over.

The formal group photograph gets taken once, on the day, and whoever missed it is absent from the only copy that exists. Here the end of the step is empty.

Same two uploads, same one sentence. She comes in at the row's scale under the same overcast light, with the shadow on the stone going the right way and the bride untouched.
Runs in the browser with free credits — nothing to install and no masking. This tool adds someone; the remove person from photo tool does the opposite when a stranger walked through your shot, and the add a loved one to a photo version handles an old album scan going into a recent picture.
From two separate pictures to one, with the cost shown before you commit.
The photograph you want to keep. Order matters: the first upload is the picture that survives, the second is the person being brought into it.
A clear, well-lit photo of their face. A passport shot, a cropped selfie or a scan from an album all work, and the more of the head that fills the frame the more there is to carry across.
Name the spot and the neighbour: on the left of the woman in the cardigan, seated in the empty chair, standing behind and between the two at the back. Precise placement is what turns a good composite into one nobody questions — and what comes back is the picture your family has been missing.
Almost every family has one: the picture everybody likes, with somebody absent from it. These are the three reasons people add person to photo.

One person took the shot, so one person is not in it. Add them back and the picture finally has everyone who was actually there that day.

A deployment, a shift, a flight that did not happen. The wedding group photo and the graduation row both get taken once, and whoever missed them stays missing for good.
The most asked version of this, and it has a page of its own — the add a loved one to a photo tool takes a scan from an old album into a picture shot years later.
What people ask before they add person to photo on Kavel.
That is exactly what this page does, and it is why it takes two uploads instead of one. The first picture is the scene that survives; the second is the person being placed into it. A tool that takes only one upload has to invent the second person from your description, which is why those results never look like anyone in particular.
Placement and light do most of the work. Say precisely where they go and who they stand next to, rather than asking for them to be added somewhere; then check three things on the result — their scale against the nearest person, the direction of their shadow against everyone else's, and the edge where their shoulder meets the background. Those three are what a viewer notices first, and they are the three this tool matches for you.
Upload the group shot, upload a photo of the person who is missing, and name the position they should take. There does not need to be a gap in the frame already — people can be placed at the end of a row, half behind a shoulder, or into a chair that was left empty.
Every account starts with free credits and the tool runs in the browser, so you can add person to photo on a real group shot today. The credit cost of each run is shown before you press generate, and a render that errors out is refunded automatically.
By uploading their face as the second picture rather than describing it, which is what makes the result recognisable. Family spot them instantly. Whatever is visible in that upload carries across — age, hair, beard, glasses, and the clothing if the shot includes it — and the face is rebuilt at the size and angle the scene needs so it sits in the picture rather than on top of it. A sharp, front-facing second photo gives the most to work with.
Manually means masking them out of one picture, dropping them into the other, then spending the real time on colour, edge feathering and a shadow that matches the light. The masking is the easy half. This tool exists for the second half, which is where hand composites fall apart.
Credits are shared across everything Kavel runs. You see the cost before each run, and a failed job is never charged.
Pay safely and securely with
Payments are processed by Stripe over an encrypted checkout. Your card details are handled by Stripe and are never stored by Kavel.
The group shot stays exactly as it was taken — the light, the faces, the day — and the person who was not standing there is standing there. Your account already has free credits on it, so the first one is on the house.