Discover the story written on your face. AI-powered 468-point facial analysis meets traditional face reading — all processed privately on your device.
Combining modern AI with traditional wisdom
Capture a selfie or upload a photo. Supports JPEG, PNG, and WebP formats for instant analysis.
MediaPipe Face Landmarker detects 468 precise facial feature points, mapping every contour with sub-millimeter accuracy.
AI classifies your face into one of five shapes — Oval, Round, Square, Heart, or Long — with confidence scores and detailed metrics.
Generates a comprehensive Mianxiang report covering Three Divisions, Five Features, aura reading, and life-stage insights.
Detects your current emotional state — calm, joyful, sad, angry, surprised, and more — as part of your aura reading.
Copy your structured analysis data for deeper AI-powered readings with any large language model.
Advanced AI, running entirely on your device
Google's MediaPipe detects 468 facial landmarks in real-time, mapping the precise geometry of your face — eyes, nose, mouth, jawline, and forehead contours.
A custom ONNX machine learning model classifies face shapes using the landmark data, providing confidence scores and detailed geometric measurements.
All computation happens on your device. No photos are uploaded, no data leaves your phone. The AI models run directly in your browser or app.
The analysis combines AI-detected facial geometry with traditional Chinese face reading principles — Three Divisions (三庭), Five Features (五官), and aura interpretation.
We built privacy into the core of everything
All analysis runs locally. Your photos never leave your device — no cloud processing, no server uploads.
We don't collect names, emails, locations, or any personal information. Period.
Zero analytics, zero ads, zero third-party SDKs. Your face is your business.
Available on iOS and Android