AI Feynman summaries + auto flashcards. Built for knowledge workers who save more than they read.
You save articles. You don't read them. The ones you read, you forget. Anti Noise fixes that with AI doing the work you'd never get around to.
Articles the average knowledge worker saves per week.
Articles actually read from that pile each week.
Articles whose ideas you still remember a month later.
Three steps. Two of them happen automatically.
URL, note, or screenshot — via Share Extension from any app. Safari, Twitter, Photos, Notes. One tap.
GPT-4o writes a simple explanation, an analogy, knowledge gaps, a real example, and a deeper question. Always 5 parts.
3-15 SM-2 spaced-repetition cards per summary. Open the app — you only see what you're about to forget.
TL;DR optimizes for "didn't have to read it." Feynman optimizes for "understood it well enough to teach it." Two opposite goals.
The same spaced-repetition algorithm that runs Anki — without the chore of making cards manually. The AI picks them.
Pomodoro-style sessions with wall-clock accuracy. Background-safe — keeps running when your phone locks.
Captures live on-device (SwiftData). Optional Firebase sync mirrors across your devices. Your data stays yours.
v1.0.1 provides AI included. Sign in with Apple or email — start capturing immediately. No OpenAI account needed.
App is bilingual from day one. Capture in either language. Summaries respect your content language.
Start free with 5 AI summaries/month. Upgrade when you want unlimited.
Follow the build, the bugs, the cost math, the launch numbers — honestly, on X.
Follow @antinoise_app →v1.0 is in App Review now. v1.0.1 (the "no API key" version) ships ~2-3 weeks after that. Subscribe above to get the App Store link the day it goes live.
v1.0 requires bringing your own OpenAI API key — friction I'm removing in v1.0.1. The waitlist is for the version where you just sign in and the app provides AI. Worth waiting two extra weeks for.
Not yet. iOS first. Mac Catalyst + browser extension are on the roadmap — but only after v1.0.1 has real users. Honest answer: subscribe and I'll tell you if the priority shifts.
Captures live on your device first (SwiftData). Optional Firebase sync mirrors across your iPhones/iPads. AI calls go through Anti Noise's backend — your captures aren't shared with anyone else, not used for training, never sold.
Normal "TL;DR" summaries optimize for skimming — you read less, remember less. The Feynman method (simple explanation, analogy, knowledge gap, example, open question) optimizes for understanding well enough to teach. Different goal, different outcome.
Not at launch. The Share Extension covers most capture use cases. If you have a large existing library and import would matter to you, reply to the launch email and I'll prioritize accordingly.
One email. The App Store link. No nonsense.