WebGL shader remaps red-green (Deutan & Protan) and blue-yellow (Tritan) confusion. Adjustable severity 0β100%.
Tap a pixel; the app speaks its plain-English name ("warm olive", "traffic red") and shows the exact hex.
OCR the live frame; text-to-speech reads it back. Type a target language to translate on the fly.
One tap narrates what's in front of you β useful when colours alone don't give enough context.
Today it runs as a web app on the phone paired to your Meta Ray-Ban / smart glasses β you hear colour names, read-aloud and scene descriptions through the glasses' speakers, and can drive it hands-free with your voice assistant. A native on-glasses build follows as Meta opens its developer SDK. It also works on any phone or tablet with a camera, with no glasses at all.
Free during the beta β no card and no account. If a paid tier arrives later it will be opt-in, clearly priced before you pay, and the core colour-correction lens stays free for the people who rely on it.
Yes. The colour lens and tap-to-name run entirely on your device β the image never leaves it. Read-text and describe-scene send a single still frame to an AI service only when you tap that button; frames are processed for the answer and not stored. No ads, no tracking, no third-party analytics.