Press

MirrorMuse AI — For journalists, editors, and writers

Thank you for considering MirrorMuse. Everything you need is on this page. If something is missing, email press@mirrormuse.ai and we will respond within four hours during UK business hours.

1. The story in one line

MirrorMuse is the honest AI beauty mirror. It tells you what works and what does not, the way a beauty editor would, not the way an influencer would.

2. The story in 100 words

MirrorMuse is a beauty AI app built in London that gives you an editorial-grade critique of your makeup. Instead of telling you that you look flawless, MirrorMuse scores your look 1 to 10 across four dimensions and tells you specifically what to fix. You can also try a curated catalogue of 60 lipstick, eyeshadow, blush, and brow shades on your own face before you commit, and you get a full beauty profile mapped to your specific skin tone, undertone, and face shape. The app launched on the App Store on 19 June 2026 and is built by Astragrid Technologies Ltd.

3. The story in 250 words

MirrorMuse is a beauty AI app launched on the App Store on 19 June 2026 by London-based Astragrid Technologies Ltd. The app was built in response to a long-running pattern in beauty technology: the "you look amazing" feedback loop that makes Instagram filters so frustrating. MirrorMuse is the opposite. Upload a selfie with makeup and MirrorMuse returns an editorial-grade scored critique, naming specifically what works and what does not, with concrete suggestions for what to change next time. The scoring is broken into four dimensions: Harmony, Technique, Choice, and Polish. There is a Try-On feature that renders a curated catalogue of 60 lipstick, eyeshadow, blush, and brow shades (in matte, satin, gloss, and other finishes) on the user's own face in under a minute, so users can stop guessing in the beauty aisle. A Face Analysis module produces a detailed beauty profile mapping skin tone, undertone, face shape, eye shape, lip fullness, and brow shape to the palettes that actually suit them. An Editorial Coach surfaces a daily tip and a curated trending feed pulled from beauty editorial. The app is free to try with a one-time trial of 3 uses per feature. Premium is GBP 12.99 per month or GBP 99 per year with a 7-day free trial. MirrorMuse never trains its models on user photos. The app works equally well across every Fitzpatrick skin tone, an explicit design constraint benchmarked before launch.

4. The story in 500 words

MirrorMuse is the first beauty AI app that admits when a look does not work. Built in London by Astragrid Technologies Ltd and launched on the App Store on 19 June 2026, MirrorMuse is positioned squarely against the flattering tone of most beauty technology. Where Instagram filters smooth and Snapchat softens, MirrorMuse tells you that your eyeliner is uneven on the left, your foundation is two shades too warm, and your lip colour is fighting your undertone, then tells you exactly what to do about it.

The product is the result of a year of work on AI critique benchmarks, particularly on the long-overlooked problem of skin tone accuracy. Most existing beauty face-analysis models were trained on predominantly Eurocentric datasets and fail on Fitzpatrick V and VI. MirrorMuse was benchmarked across all six Fitzpatrick ranges and refused to ship until the accuracy was consistent across them.

Three features anchor the app. Face Analysis gives the user a structured beauty profile mapping skin tone, undertone, face shape, eye shape, lip fullness, and brow shape to recommended palettes. Critique scores the user's current makeup 1 to 10 across four locked editorial dimensions (Harmony, Technique, Choice, Polish) with specific strengths, specific weaknesses, and concrete suggested fixes. Try-On generates a photographic preview of any lipstick, eyeshadow, blush, or brow shade on the user's own face in under a minute, and lets them save their favourites to a private library.

An Editorial Coach surfaces a curated daily tip and a rotating trending feed of beauty editorial from across the web, replacing the influencer recommendation algorithm with editorial taste.

The free tier offers a one-time trial of 3 uses per feature, after which Premium unlocks unlimited access at GBP 12.99 per month or GBP 99 per year with a 7-day free trial.

MirrorMuse never trains its AI models on user photos, never sells user data, never shows ads, and never tracks users across other apps. The full privacy policy at mirrormuse.ai/privacy documents this position in detail. The company offers a published delete-account flow at mirrormuse.ai/delete-account and treats this commitment as a core brand value, not a regulatory checkbox.

MirrorMuse is available globally on the App Store, with the Android version planned for July 2026 following Google Play's mandatory closed-testing window for new developer accounts.

5. Quick facts

6. Brand assets

Hi-resolution brand assets for use in print, web, and broadcast. All assets are free to use in editorial coverage of MirrorMuse with attribution to Astragrid Technologies Ltd.

• App icon (1024 x 1024 PNG) — download
• App icon (512 x 512 PNG) — download
• Wordmark — render in Bodoni Moda 600. No bitmap wordmark provided; the typography is the wordmark.
• App screenshots (4 PNG, 1290 x 2796 each) — download ZIP
• Brand video (60s, MP4 1920 x 1080) — available on request
• Editorial lifestyle imagery — available on request

7. Quote-worthy lines

For use in articles, attributable to MirrorMuse:

"Most beauty apps tell you you are flawless. MirrorMuse tells you the truth, kindly. There is a difference between a beauty critic and a hype cheerleader, and the industry needed one of those."

"The fact that mainstream face-analysis AI silently fails on Fitzpatrick V and VI is the dirty secret of the beauty tech industry. MirrorMuse refused to ship until the accuracy was consistent across all six skin tones."

"Honest is the new flattering. Beauty consumers are tired of being told they look amazing when they know they could look better."

8. Story angles

• The honesty trend in consumer AI — why anti-flattery is a positioning that resonates in 2026
• Skin tone equity in face-analysis AI — benchmarking accuracy across Fitzpatrick I to VI
• Beauty tech beyond Instagram filters — what AI in beauty can actually be useful for
• Try-On vs counter samples — how AI try-on changes beauty buying behaviour
• Privacy as a beauty product feature — no training on user photos, no data sale, no ads

9. Contact

Press inquiries: press@mirrormuse.ai

Asset requests, additional imagery, or extended product information: press@mirrormuse.ai

General enquiries: hello@mirrormuse.ai

We respond within four hours during UK business hours.