Copy a YouTube link → open app → it auto-detects from clipboard, pulls metadata + transcript, and generates a structured summary (key points + narrative) in seconds. No inputs, no pasting, no prompting. Supports 60+ languages for summaries, the video's language doesn't matter.
Why I built:
I had an endless queue of educational videos but no time to sit through rambling or clickbait. ChatGPT flow = copy → open → paste → prompt → pray it formats correctly. Existing summarizers felt feature-bloated or unreliable. I wanted something super minimalist: “clipboard → value” on repeat.
Decisions:
Stack — React Native 0.81 (New Arch) + TypeScript
Clipboard trick
— Monitors clipboard on app start/resume — no manual paste
Caching
— Stores transcripts + summaries in AsyncStorage with metadata
— fast image library prevents repetitive refetching
APIs
— YouTube metadata
— Third-party transcript provider (exploring Whisper)
— OpenAI for summarization
Output
— Structured prompts → consistent format across 60+ languages
— Easter egg: philosophical analysis for songs/poems
Time-saved calculation
— Uses actual video length (30% of cases)
— Fallback = estimated from average speech speed
Main headache:
Grandfathering legacy users without breaking App Review was harder than I expected. The first version was paid upfront ($14.99 or leave). Got some sales in the first two weeks, then traffic died. After some internal fighting, I switched to free-to-try with IAP. By then I already had thousands of users I’d promised free access forever, plus a dozen who actually paid. I wanted to keep my word and never show them a paywall.
RevenueCat’s originalAppVersion looked perfect… until I actually used it.. Turns out it’s not the marketing version (1.0, 2.3), but the build number (the variable name originalAppVersion is so misleading…). And Apple Review/TestFlight always report it as 1.0, so reviewers looked like legacy users and never saw the paywall (5 rejections). Everything worked for me, totally broken for them.
My mistake: I auto-entitled anyone with 1.0. First hacky fix: show a review-only paywall. Proper fix: if originalAppVersion === "1.0" (I never had such a build number anyways), treat as non-legacy. Days of digging and testing across devices/envs… finally approved.
Current state:
3800+ users and 12 sales within 2 weeks post-launch, 4.9-star rating globally. Now multi-tier subscription after 3 free summaries due to ongoing costs and long-term value.
Limitations/Challenges:
— iOS only for now. Android clipboard monitoring is restricted—would need a different UX.
— Some countries need VPN due to API restrictions; I notify users on app start.
— Dependency on third-party transcript API (evaluating OpenAI-based solution).
— YouTube's testing AI conversational tools, but it's not quite the same thing, sandboxed only for some users and unlikely to roll out widely as it cuts ad their revenue from watching.
Happy to answer questions about implementation or share code snippets if helpful.
https://apps.apple.com/us/app/esse-ai-video-summarizer/id675...