Take one photo and find out. An AI model compares your face against 321 dog breeds and returns the one you most resemble, along with the specific features that gave it away — hair, build, the shape of your face, the expression you happen to be wearing. It is free, it needs no account, and your photo is never stored.
It is not a filter. Nothing is drawn on your photo and nothing is generated: the model reads the picture once, writes its verdict, and the image is discarded.
You take a selfie — on a phone the button opens the camera, on a computer it opens your files. The photo is sent to our server once, passed to an AI model along with the breed catalogue, and discarded. Nothing is written to a database, nothing is saved to storage, and the image is not logged.
The model is told to work only from features a dog also has: hair length, texture and colour, how a fringe falls, build and proportion, the geometry of the face, and the expression someone wears. It is explicitly forbidden from commenting on attractiveness, weight, health, age, ethnicity or identity. The joke is the resemblance, never the person.
No, and it would be a bad way to. Looking like a Border Collie says nothing about whether you have two hours a day to exercise one. If you are genuinely choosing a dog, the useful pages on this site are the breed guides and the ranked lists — they cover energy, grooming, shedding, tolerance for being alone, and the breeds people commonly pick and regret.
It is sent once, read, and thrown away. DoggyQ does not store your photo, does not save it to a database or file storage, and does not use it to train anything. It exists for the few seconds the match takes and then it is gone.
Yes — free, with no account, no email and no payment. Nothing is held back behind a signup.
It runs on Claude, Anthropic's AI model, reading the photo directly — there is no filter, no template and no pre-set list of celebrity lookalikes. The model is given the photo and the full list of 321 breeds, and has to justify its pick by naming the features you share, which is why the answer explains itself rather than just asserting a breed.
No. A filter puts dog ears on your photo; this does not change your picture at all. It reads the photo once and writes a verdict about which breed your face rhymes with, then throws the photo away. Nothing is drawn, morphed or generated.
No, and it will decline if you try. It matches a human face to a breed, not a dog to its breed. If you want to identify a dog, this is the wrong tool — but the breed guides here will tell you what each breed actually looks like.
An AI model looks at the photo and compares it against 321 breeds on the features a dog also has: hair length, texture and colour, build and proportion, the shape of the face, and the expression. It returns the closest breed, the specific features you share, and a runner-up.
No. The model is instructed never to comment on attractiveness, weight, health, age, ethnicity or identity, and only to work from features that translate to a dog — hair, build, face shape, expression. The joke is the resemblance, never the person.
It is a resemblance game, not a measurement. There is no correct answer to "which dog do you look like", so treat the result as a playful read rather than a fact. If you are actually choosing a dog, the breed guides and ranked lists on this site are the useful part.
Only use photos of yourself, or of an adult who has agreed to it. The matcher declines images where the subject appears to be a child.
No. It runs in the browser. On a phone the button opens the camera directly; on a computer it opens your file picker.
Every breed in the matcher has a full guide behind it, built from the same trait data.