AI color analysis
Zivya finds your real undertone from one selfie — across the full range of skin tones, including the medium, deep and rich complexions other apps get wrong.
Your first analysis is complete and free. No paywall over your result.
An illustration of how results are built. The real analysis reads your undertone from your own photo.
Why Zivya exists
The complaints repeat and they are documented: results that lighten the skin, features quietly narrowed, foundation shades recommended that do not exist past a certain depth. We built Zivya around that failure.
The analysis describes your tone exactly as it appears. No shifting toward a lighter value, ever.
Your nose, lips and face shape are described as they are — not reduced, not corrected.
Six depth levels, with deep and rich as real categories, not an afterthought at the end of a scale.
Bad light, a filter, no face — Zivya tells you, instead of guessing a result that looks confident and is wrong.
How it works
Natural light, no filters. The app guides your framing so the color reads true.
Warm, cool, neutral or olive — with your depth, your color season and your face shape.
Real color swatches that complement you, and the ones that flatter you less.
Morning, night and makeup steps matched to shades from Walmart, Sally and Ulta.
The difference, plainly
Questions
Zivya is in development for iPhone and Android. Email hello@zivya.app and we will tell you the day it ships.
Yes, and you see all of it. Hiding the first result behind a paywall is the most common complaint about apps in this category, so we do not do it. After that, the free plan includes one analysis per month.
It is stored privately, never public, and served only through expiring signed links. Photos are deleted automatically after 30 days, and you can erase your account and everything in it at any time from Settings.
That is the entire reason the product exists. The analysis is validated specifically against medium, tan, deep and rich complexions before anything ships.
Yes. The app is in English by default, and Spanish is available from onboarding and Settings.