About

What PawfectPix does

Upload one photo of your pet and get back a stylised portrait in the art style you pick — plus, if you want it, a short and affectionate personality roast. No studio, no sitting, no waiting for a photographer.

Create a portrait

How it works

  1. Upload a photo

    JPG, PNG or WEBP, up to 10 MB. A sharp, evenly lit photo with your pet's face clearly visible gives the best result; a blurry or dark snapshot still works, but the likeness will be looser.

  2. Pick a style and a size

    Six styles — watercolour, royal, cute 3D cartoon, superhero, cosy and minimal studio. Choose 1K, 2K or 4K depending on whether you want a quick preview or a file big enough to print.

  3. Generate

    One portrait costs 10 credits and usually takes 20 to 60 seconds; the page stays usable while it runs. If the image service fails or refuses the photo, the credits go straight back and that attempt is not charged.

  4. Download and share

    The finished portrait is watermark-free and yours to download. The optional personality roast appears beside it as a short piece of text you can copy.

What happens behind the scenes

A style is a prompt, not a filter

Each style is a carefully written instruction set given to an image model alongside your photo. That is why the result looks painted rather than pasted over, and why changing the style changes the whole image rather than just its colours.

Two models, named

Google's Gemini Nano Banana 2.1 keeps your pet's actual face closest and outputs 1K, 2K or 4K; OpenAI's GPT Image 2.5 paints more boldly and outputs 1K only. The picker only offers resolutions the selected model really supports, because sending 4K to a 1K model produces a silent downgrade.

Your photo is used to make this portrait only

It is sent to the image service for that one request. We do not use it to train models, we do not publish it, and we do not keep a copy of the uploaded original once the portrait has been generated.

Portraits are stored so a refresh does not lose them

Finished images are kept in our own object storage and served from our own domain, so a link you keep will still work. What you download is never the provider's temporary URL.

The roast is written separately

The personality text is produced by a text model looking at the same photo. It runs in parallel with the image; if it fails, the portrait is still delivered and nothing extra is charged.

Credits, not a subscription

You buy a pack once and spend it whenever you like. Credits never expire, nothing renews, and a failed generation is refunded automatically. The exact rules are in the Terms of Service.

The models we actually use

We name them instead of writing "an AI image model", because which model runs is the reason one portrait keeps your dog's face and another turns it into a painting.

Model access is supplied through Grsai, an API provider; the models themselves belong to Google and OpenAI. If we change or add a model, the create page, this page and the Privacy Policy are updated first.

What it does not do

Where your data goes

Your photo travels to the image generation service for the request that makes the portrait, and the finished image is stored in our own object storage. We do not sell your data, we do not use photos to train models, and we do not keep the uploaded original after generation. The Privacy Policy sets out exactly what is stored and for how long.