About Mirri
A skin scoring app that runs on a single selfie.
Why we built this
Going to a dermatologist every time you're curious about your skin is a hassle, and the bathroom mirror doesn't tell you much. We wanted something casual you could check now and then. One selfie, a score against people in your age and gender bracket, and a hint about what to try next — that's the bar we set.
How to use it
Take a selfie
Bare face, front-facing, bright light. Stick to those three and the results stay consistent. You can shoot with makeup on, but the score will lean conservative.
Let the AI score it
Seven categories, scored relative to people your age and gender. Being in your 40s doesn't auto-deduct points — average for your peers is just average.
Read the score and tips
You'll see scores out of 100, an estimated skin age, and a few care tips. It's enough to know what to focus on until your next check.
The 7 things we look at
Moisture
How smooth and lightly luminous the surface looks. When moisture is low you'll see roughness, flaking, and that tight feeling.
Pores
How visible the pores around the nose and cheeks are. Genetics, sebum, and age all play in — fully erasing them isn't realistic, but smoothing the look is.
Pigmentation
Spots, uneven tone, dark circles. UV is the biggest driver, so consistent sunscreen does most of the heavy lifting.
Elasticity
Sagging cheeks and deepening nasolabial folds. Collagen drops with age — points only come off when you're below your peer average.
Wrinkles
Around the eyes, forehead, and nasolabial fold. Facial habits and UV matter most; hydration helps too.
Redness
Redness around the cheeks and nose. Sensitive skin or reactivity to temperature shifts can pull this score down.
Oiliness
How shiny the T-zone is compared to the cheeks. T-zone-only shine usually means combination; all-over shine is closer to oily.
What's skin age?
Not your actual age — it's the age your current skin condition most resembles, estimated from your seven scores. Take care of it and the number reads younger; neglect it and it creeps up. Don't read too much into a single result; it's most useful as a trend over time.
Under the hood
Computer vision looks at the photo first to pick up texture, color, and patterns. A language model then turns those signals into something readable. Two models, two jobs.
Heads up
Mirri isn't a medical service. The results are for reference and don't replace a dermatologist. If something feels off, or symptoms keep coming back, please see a doctor.
Want to try it?
First analysis is free after signup.
Try it free