YOUR NEXT LOOK

Pick your inspiration.

A YouTube tutorial or a photo of the look. Add your palette and Cue uses your own shades.

Checking the analysis service…

Photos you add go through our server to Google Gemini to plan the steps and aren’t saved. Location data is removed first.

Cue Mirror

Your look, step by step

Turn on your camera to place each step directly on your face.

Camera frames stay on this device
✧Follow your cheekbones
CAMERA OFFTurn on your camera for face-mapped guidanceYour camera stays on your device
Overlay strength22%
THE TECHNIQUE

Peachy cheeks

    ✧ MIRROR NOTE

    Made to move with you

    Turn on your camera to anchor the guide to your own face.

    Placement is approximate. Left and right callouts refer to the sides of the displayed mirror. The live camera is mirrored. Guides follow face landmarks and never identify your face. Check my work compares your left and right sides on this device; frames aren’t saved or sent.
    YOUR ACCOUNT

    Sign in to analyze.

    Use the same Google account you sign in with across your Curated apps. You’ll come straight back here afterwards.
    DISCOVER

    More tutorials

    CUE MIRROR

    Your look

    Your captured look

    Photos stay on your device unless you share them.

    PREVIEW MAKEUP

    Your look, previewed

    PRODUCT

    Search on Google ↗
    VOICE

    Talk to Cue

    Ask anything about this tutorial — the steps, the products, what to do next — or say “next step” and Cue moves you along, hands-free.

    While the mic is on, your voice goes to OpenAI to power the conversation; Cue Mirror doesn’t record or save it. Cue stops after 2 quiet minutes, or tap the mic.

    PROFILE
    TUTORIAL
    Open on YouTube ↗
    HOW IT WORKS

    Your tutorial, translated to you.

    1. Bring a public tutorial

      Paste a public YouTube link or browse makeup videos. The server reads the video’s YouTube preview frames and captions, and Google Gemini turns them into steps (sign-in required). The featured look is saved, so it needs no analysis. Private and unlisted videos aren’t supported.

    2. Or recreate a look from a photo

      On Create, add a photo of a look (and, if you like, photos of your own palette). They go through our server to Google Gemini, which writes the steps; Cue Mirror doesn’t save them, and location data is removed before sending. AI can misread a photo, so treat the steps as an approximation.

    3. Review your sequence

      AI-generated steps include instructions, products when identifiable, and estimated source timestamps. Check them against the video. AI can miss or misunderstand details.

    4. Practice in your mirror

      On-device face landmarks anchor short instructions and translucent guides to your cheeks, eyes, lips, and brows. Your live camera is never recorded or sent to our server or Gemini. If you tap “Preview makeup”, one selfie and the tutorial’s YouTube images (or, for a look from a photo, your inspiration photo) are sent through our server to OpenAI, which creates an AI image of you wearing the look; neither is saved by Cue Mirror. Photos you capture stay on your device unless you share them.

    YOUR CAMERA STAYS LOCAL

    The browser downloads a face-tracking model and processes camera frames on your device. When you tap Check my work, it also reads the colours of the camera frame around your eyes, brows and cheeks on this device to compare your two sides; those frames and measurements are never saved or sent. Google Gemini receives the tutorial’s preview frames and captions when you analyze. When you talk to Cue, your voice goes to OpenAI only while the mic is on. OpenAI receives one selfie, with the tutorial’s YouTube images, only when you tap Preview makeup and confirm. YouTube receives normal requests when you browse, see thumbnails, or open the tutorial. Your look and progress stay in this tab.

    Guides suggest approximate placement, not a guaranteed match. Lighting, complexion, facial structure, and products change the result. Keep makeup away from the eye surface and follow product directions. Spoken guidance uses your browser’s speech service; its processing depends on your device.