AI background changer

To change the background of a photo with AI, upload it, pick a scene from the list or describe one in a sentence, and press Change the background. The model keeps the person or product as photographed and draws the new scene around them, with the light and shadows matched. Make it with any DreamBench plan.

No. Every run needs a DreamBench plan; you create an account on the way, and the button shows the price in credits before you press it, which depends on the size of the photo. A run that fails is refunded. The plan covers every tool on the site and the video models.

No, you don't need to remove the original background first. Upload the photo as it is and pick or describe the new scene: one run replaces the background around the subject, keeping the subject and drawing the new scene in the same step. A cutout works too; its see-through pixels are read as white.

Hair usually comes through well. Glass, chrome and polished surfaces reflect the room they were photographed in, and the model redraws those reflections for the new scene, so check them closely before you use the picture.

Yes. Run the same photo again with another scene, or with a new description: each run is a new picture made from the same upload, so you can compare a studio, a surface and a place side by side. Each run takes one photo; there is no batch upload and no API.

For one flat color, use the background color changer: it cuts the subject out pixel for pixel and fills the background with a swatch or a hex value instead of drawing a scene. For a cutout on transparency, the background remover. This tool's result is one flat JPEG with the subject drawn into the scene, with no layers and no transparency.

It sends your photo and one instruction to an image model that edits from references. The instruction names the new scene and tells the model to keep the subject exactly as it is, matching the light and shadows to the new background.

The instruction locks the subject: the same person or object, pose, proportions, colors, clothing and details. Small differences can still appear on fine text or very detailed patterns, so check labels and logos before you publish.

Yes. Choose Describe your own and write the scene in a sentence, such as a rooftop at night with the city lights behind. The words become part of the instruction, and the light on your subject is matched to the scene, so say where the light comes from if it matters.

Removing a background cuts the subject out and leaves it on transparency or plain white, pixel for pixel. Changing a background draws a new scene around the subject and relights it to match, so the whole picture is rendered again. Use the remover for a cutout and the changer for a place.

Partly. Canva can remove the background and let you lay the cutout over another photo, and an iPhone can lift a subject out of a picture. Neither relights the subject for its new scene, which is what this tool adds, so a pasted subject often looks pasted. This page runs in the browser on a phone, a tablet or a computer, with nothing to install.

Three things give it away: light from the wrong side, a subject cut off by the frame, and reflections of the old room in glass or chrome. Pick a scene whose light matches the photo, upload the whole subject, and say where the light comes from in your description.

There is no visible mark, and DreamBench adds no watermark of its own. Every image the model makes carries SynthID, Google's invisible watermark that identifies it as AI-generated; it doesn't show in the picture and doesn't change how it looks.

The result is a JPEG in your photo's proportions, rendered at the model's 2K tier when the upload has more than about 1.5 megapixels and at its 1K tier otherwise. A photo larger than 2048 pixels on its long edge is scaled to 2048 before the run, so a 4K photo works but comes back smaller than it went in.

No. It changes the background of one still photo at a time. To make a moving picture, change the background of a still here, then press Animate this scene: the composer opens with it as a reference, and DreamBench's video models make a new clip from it.

A run starts at 7 credits from your DreamBench plan, depending on the photo, and once you have a plan the button shows each run's price before you press it. Starter, $9.99 a month, includes 1,100 credits a month, at least 122 runs of the AI background changer, plus AI video. A run that fails is refunded.

Yes: a free account and a DreamBench plan. The tools are part of every plan, from Starter at $9.99 a month.

A photo up to 40 megabytes: JPEG, PNG, WebP, HEIC or another format your browser opens. A photo straight from a phone is fine, an iPhone's HEIC included: the page scales a large picture down to 2048 pixels on its long edge before the run, so the result is never larger than the page says.

Yes. Under the DreamBench Terms you may use what a tool makes for personal and commercial purposes, including advertising, monetized content and work for clients, with no separate commercial-use upgrade. The Terms don't guarantee copyright in a result or clear another person's rights in your upload, so only use photos you have the rights to.

No. DreamBench doesn't use your prompts, uploads or videos to train or fine-tune AI models, and any future training program would be opt-in.

It stays in your browser until you make a result. Then it is used to make that result and nothing else, and removed from the tool's storage after seven days along with the result. A result you keep in your library stays as long as you keep it.