Hey Product Hunt 👋 — Sean here, indie maker behind Beast Teleprompter. Quick story on why this exists: I spent 4 years as an AI product manager working on LLM applications. The whole time I kept watching people — colleagues, friends, my family — freeze up on camera. Not because they didn't know what to say, but because the teleprompter was actively working against them. Fixed scroll speeds force you to read at the prompter's pace, not yours. Pause for emphasis → text leaves you behind. Speed up → you're chasing it. The tool that was supposed to help became the source of friction. And on top of that, the workflow itself is fractured: prompt in one app, record in a separate camera app, then move the file out and deal with a watermark. Three apps and a file shuffle for one talking-head video. So when I went indie I built the version I wanted: a teleprompter that listens, and records video in the same window. Beast uses voice tracking to match its scroll speed to your delivery in real time — you can pause to think, speed up when you're confident, repeat a line, improvise, and the script just keeps up. Recording is built in, exports are 1080p with no watermark on the free tier. One app, one window, talking-head video done. A note on the "on device" line in the tagline — I get asked about this a lot. Beast runs the speech recognition fully on your device, no audio is uploaded, no transcript leaves your phone, nothing routes through a server. Two reasons it matters: 1. Latency. Cloud ASR adds a network round-trip on every chunk of speech. For a teleprompter that has to track your real-time delivery, those few hundred ms are the difference between "follows you" and "lags behind you." 2. Privacy. Most creators read scripts that contain unpublished material — campaign briefs, course content, talking points for an interview. Sending all of that to someone else's server because you wanted a teleprompter to scroll faster is a bad trade. A few things I deliberately put in the free tier (because creator tools paywalling basics is a pet peeve): • AI voice tracking, no time limit • 1080p recording, no watermark • All the viewing modes (mirror, flip, portrait, landscape) • Recording studio with built-in prompter Pro adds the things that genuinely cost more to build or run: • Floating overlay that works on top of TikTok, IG cam, etc. • 4K recording • Unlimited scripts • Cross-device sync (iPhone ↔ iPad ↔ Mac) One last thing on where Beast sits in the current AI wave. So much of building right now is either making products for agents to call, or going end-to-end synthetic — AI avatars, virtual presenters, fully generated video. Both are interesting bets. Mine is different: I think messy, real humans still deserve better tools to express their messy, real selves. Probably more than ever. Beast doesn't replace you and doesn't speak for you. It listens while you talk — badly or beautifully — and tries to disappear. Happy to answer anything — the on-device speech stack, the indie journey, why I picked specific design decisions, or what's next on the roadmap. Honest feedback wanted. If you've ever had to re-record a take because the prompter wouldn't slow down, this one's for you 🎬

Beast Teleprompter

the teleprompter that follows your voice — on device

the teleprompter that follows your voice — on device