Press

Ten pieces
is a wardrobe.

Cinch is built on one observation: nobody finishes cataloguing their closet. Every app in this category asks you to photograph a hundred garments before it does anything for you. Most people quit around the twelfth.

What it is

Cinch starts at ten. You photograph three tops, two bottoms, a dress, two pairs of shoes, a coat and a bag — and it tells you that you own fourteen outfits, because three times two times two, plus one times two, is fourteen. That number is computed from your actual closet, not claimed.

From there it dresses you every morning: for the weather where you actually are, what is on your calendar, and what you reach for. And when there is an occasion — a wedding, a dinner, a trip — the people going with you can see two options from your real closet and help you choose.

The part that is unusual

Outfit generation is not an API call. It is a deterministic engine that runs on the phone: perceptual color math in OKLCH, a beam search over your wardrobe, and a linear model that learns from every reaction. It works with no signal, it costs nothing per outfit, and every recommendation decomposes into a reason you can argue with.

A language model never chooses an outfit. It reads a photograph of a garment when you add it, turns a spoken walkthrough of a closet into a list of pieces, and writes one sentence about a look the engine has already chosen. When you describe a piece to find a shop’s own picture, Cinch searches Datafeedr’s licensed catalog instead; no AI model reads that description. The choosing is the engine’s, on the phone.

Every outfit is assembled from pieces you already own. Nothing is imagined, and nothing is a piece you would have to go and buy first.

What it deliberately does not do

• There is no store. No catalog to browse and no “complete the look”. Every outfit is assembled from clothes already in the closet. Cinch links out to a shop in exactly two places — when it can name a gap the closet actually has, and when one user is looking at a piece in another user’s closet — and those links can earn a commission, which is disclosed on the link itself.
• No advertising, and no tracking across other apps.
• No public feed and no strangers. Friends are people you invited. An outfit is voted on by the people actually going, and only ever against that person’s own other option — never ranked against anyone else.

Fact sheet

Name

Cinch

Category

Lifestyle · iPhone

Price

Free · Standard $0.99/mo or $9.99/yr · Pro $2.99/mo or $29.99/yr (US, from 11 September 2026)

Free tier

Ten pieces, full styling, cost-per-wear and wear tracking

Trial

None — the free tier is the trial: ten pieces and a first outfit before any purchase

Built with

React Native and Supabase, with an on-device styling engine

Revenue

Subscription, plus possible affiliate commission on the two surfaces where Cinch links to a shop. A paid link is labeled on the link itself. No advertising, and no payment for placing any particular product.

Founder

Nelson Saenz

Press contact

hello@cinchstyle.com

Website

cinchstyle.com

Imagery

Free to use in coverage of Cinch. Full-resolution files on request.

Accessibility

All body text meets WCAG AA contrast; one mid-tone in the palette was overridden during design specifically because it failed at 3.5:1. Progress is always stated as a count — “four of ten” — never a bar, which keeps it legible to screen readers and to anyone who finds proportional indicators hard to parse. Removing a garment permanently is a separate, confirmed action; archiving it is reversible.

Brand

• Name: Cinch, always capitalised, never “the Cinch app”.
• Type: Instrument Serif for display, Archivo for text.
• Color: warm paper #FBFAF6, ink #1C1A16, a single oxblood accent #6E2B2B.
• Describe it as a wardrobe app. Not an AI stylist, not a shopping app, not a social network.