Real age, perceived age and biological age
Chronological age is the time since you were born. It is the only one of the three that is a fact, and no photo can measure it. Perceived age is how old you look to someone else, and it changes with light, expression, grooming and the camera as much as with the years. Biological age is an attempt to describe how a body is ageing, using measurements such as blood tests, and it is studied in medicine.
An AI age estimator like this one learns perceived age. It was trained on photos labelled with ages, and it outputs the age that faces like the one in your photo were most often labelled with. That makes it a perceived-age machine: it answers "how old does this face look in this photo?", never "how old is this person?" or "how healthy are they?".
What research says, in general terms
Perceived age has been studied for a long time, both with human raters and, more recently, with face-analysis models. Researchers have looked at how closely perceived age tracks real age, which visible features drive it, and whether it relates to health. Some medical research groups have built models that estimate a form of biological age from face photos for use in clinical studies.
Those are research systems trained and tested for specific purposes. None of their findings applies to this tool, which is a small open-source model built for curiosity, and nothing it shows should be read as a health signal.
Factors that shift your face age
- Lighting: side and overhead light deepen shadows in every line and hollow, which reads as older.
- Camera distance: a phone held close distorts the nose and cheeks.
- Expression: a broad smile adds lines around the eyes; a tired or squinting face adds years.
- Skin texture: filters and beauty modes remove it and usually pull the estimate younger.
- Facial hair, glasses and hats: they hide the brow, forehead and jaw cues the model uses.
- Makeup: depending on coverage and contour it can move the estimate either way.
- Sleep and dark circles: puffiness and shadow under the eyes are read as age; the Dark circles analysis looks at that area on its own.
Try it: compare several photos of yourself
The clearest way to see the gap between face age and real age is to hold everything else still and change one thing. Take two photos a minute apart, one facing a window and one with the window to your side, and put them in Compare. Then try with and without glasses, or smiling and neutral. The differences you see are differences in the photo, not in you.
If you type your real age, the tool shows how far each guess is from it. Share card turns the result into a 1080×1350 image made on your device, with the estimate, its range, the difference from your real age and a small line saying it is an estimate, not verification. You can show the face crop, blur it or hide it before you download or share.
Why a single number is misleading
A single face age figure hides how much the model wavers. The stability check runs the model on the original face, a mirrored copy and two slight crops and shows the spread. A wide spread means the photo is hard for the model; a narrow one means it was consistent, which is still not the same as being right. Age models are also less reliable for children, older adults and some skin tones, so for those faces the gap between face age and real age says even less.