Portraits and team photos
Turn a snapshot taken in an office, a cafe or a hotel room into a portrait whose background reads as a soft, deliberate setting rather than a distraction.
IMG AI AI Blur Background keeps the subject of a photo sharp and softens everything behind it. Upload a picture, choose how strong the blur should be, and download the result as a PNG. The subject is found by AI; the blur itself is applied on your device, so you can adjust it before you save.
Your blurred photo appears here
Upload a photo, set the blur strength, then press Blur background. The subject stays sharp while everything behind it softens.
5 credits per photo · 0 remaining
Definition
Output
PNG at source size
Cost
5 credits per photo
Blur
Adjustable before download
An AI blur background tool separates the subject of a photo from the scene behind it and then softens that scene, so the eye lands on the subject. It is the effect a wide-aperture lens produces in camera: a face, a product or a pet stays crisp while the room, street or skyline falls out of focus.
IMG AI finds the subject with an AI matting model, then blurs the original photo and draws the sharp subject back on top. Because the blur runs in the browser, the strength is a setting you can change as often as you like before you download, instead of a fixed result you have to pay to redo.
The subject keeps the pixels from your own photo. Nothing is redrawn, recoloured or restyled, which is what separates this from a generative edit that rebuilds the whole frame and can drift away from the person or product you photographed.
What people use it for
Turn a snapshot taken in an office, a cafe or a hotel room into a portrait whose background reads as a soft, deliberate setting rather than a distraction.
A product shot on a kitchen counter or a desk keeps the surface as context while the room behind it recedes, so the item holds attention in a marketplace gallery.
Direct attention to a room or a single piece of furniture by softening a busy view through the window or a cluttered corridor beyond the doorway.
Give a phone photo the depth of a dedicated camera by keeping the person sharp and letting the street, park or venue blur behind them.
Portraits of a dog or a cat read better when the garden, sofa or floor behind them softens and the animal becomes the clear subject.
Set a person or an object against a calm, out-of-focus field so titles and captions placed nearby stay readable on a slide or a video thumbnail.
How to use it
The tool sits at the top of this page. Nothing is charged until you press the button, and the price is printed on the button before you do.
Drag a JPG, PNG or WebP into the panel above, up to 5 MB. A phone photo is a normal starting point; the subject should be reasonably clear and separate from the background.
The subject is found first. Then drag the strength slider or pick a preset, from a subtle separation to a deep, dreamy background, until the result looks right to you.
Drag the before-and-after handle to check the edges, then download a PNG at the size of your original. Every run is saved to My Creations as well.
Cases
Three before-and-after pairs, which is what this tool is tuned for: people, products and pets on a background that should recede.
Illustration note: the six images below are concept renders that show before-and-after states, not screenshots of a recorded run. They illustrate the effect the tool applies; your own output will differ with your subject, its edges and the amount of blur you choose.
Before
AfterCase 01
The person stays sharp while the street behind them loses its detail and colour separation, giving a phone portrait the fall-off of a wider aperture.
Before
AfterCase 02
The item keeps its edges, label and contact shadow while the kitchen behind it becomes a soft field of colour, so the product is the only thing to read.
Before
AfterCase 03
Fur edges are the hard case for any matting model; the animal is kept from the original while the foliage behind it softens into a green wash.
Capabilities
The blur is a local effect, not a new generation. That keeps the subject faithful and keeps the cost to a single run.
After the subject is found, the strength slider and the four presets re-render the result in the browser. Adjusting the blur does not send your photo again and does not spend another credit.
The subject is composited from your original image rather than regenerated, so the person, product or pet is not redrawn, restyled or beautified.
The final composite is assembled on your device. There is nothing to install, and the blur is previewed live before you download.
The download keeps the width and height of the uploaded photo, so a portrait crop stays a portrait crop and a square listing image stays square.
Compare
Portrait mode needs a live camera, a manual blur needs editing skill, and a cut-out tool gives you a layer rather than a finished photo. Here is how the four approaches differ.
| IMG AI AI Blur Background | Phone portrait mode | Manual blur in an editor | Cut-out tool + separate editor | |
|---|---|---|---|---|
| What you get back | A finished photo with the subject sharp and the background blurred, ready to download as a PNG | A depth-effect photo taken in the moment, with the device deciding the edges | Whatever you build, mask by mask, in an editing app | A transparent cut-out that you still have to place and blur yourself |
| Works on a photo you already have | Yes — upload any JPG, PNG or WebP | No — the effect is captured as the photo is taken | Yes | Yes, after you cut the subject out |
| Strength adjustable after the fact | Yes — the slider re-renders the blur before you download | No — baked into the captured file | Yes, with manual re-editing | Only by redoing the composite |
| Subject fidelity | The subject is composited from your original pixels | The scene is captured, not redrawn | The subject is untouched by definition | Depends on the matting quality |
| Hard edges (hair, fur, glass) | Handled by the matting model, and worth checking on the compare slider | Usually handled in camera, with occasional artefacts | Requires manual channel and mask work to look right | Often the weakest point and usually needs repair |
| Runs where | In the browser; the blur is applied on your device | On the phone camera, at capture time | In installed desktop software | In whichever online tool you choose |
| Skill needed | Upload, set the strength, download | Point and shoot | Editing experience | Two tools and a composite step |
| What it costs here | 5 credits per photo, from the same balance as every other IMG AI tool | Included with the phone | A licence plus your time | Varies by tool |
Credits and output
Each background blur costs 5 credits and returns one PNG at the size of the photo you uploaded. Credits come from your IMG AI plan, and every tool here draws on the same balance.
The price is on the button before anything is spent, and your balance is checked before the run starts. A run the provider reports as failed returns its credits. A run you cancel, and a run that finishes but returns nothing readable, keep their charge — both are visible in My Creations.
One press is one charge. Pressing twice, losing the response on a bad connection or reloading the page all rejoin the same run instead of starting a second one.
FAQ
No. The tool separates the subject from the scene and then composites the subject back on top of the blurred photo, so the subject comes from your original pixels rather than being regenerated. Edges are the one exception: matting is an estimate, so check the boundary on the compare slider before you publish the image.
JPG, PNG or WebP up to 5 MB. You get one PNG in the same size as your upload, so a portrait crop stays a portrait crop and a square listing image stays square. A larger file is rejected at upload rather than sent to the model, which keeps the run predictable.
Yes. After the subject is found, drag the strength slider or pick one of the four presets, from Subtle to Deep. The blur is applied in your browser and re-renders as you move the slider, so you can compare a few strengths and download the one you like without another credit being spent.
Yes. The segmentation is not limited to faces. Products, pets and objects work when the subject is reasonably clear and separated from the background. Delicate boundaries such as loose fur, whiskers and transparent packaging are the cases where a fringe is most likely, so inspect the edge before using the result.
A run is charged once and tied to the task it created. Reloading the page or coming back later rejoins that same task instead of starting a new one, and a run the provider reports as failed returns its credits. A run you cancel yourself keeps its charge, which is why the compare view opens before you download.
You need to be signed in, because a run spends credits from your balance. Uploads are stored on IMG AI storage so the segmentation service can read them, and the tool only accepts images uploaded through this site. Your result and the original are both listed in My Creations.
The tool is at the top of this page. Upload a photo, set the blur strength, compare the edges, and download a PNG with the subject kept sharp.