I built it out of a simple frustration: almost every "free" transcription tool wants to upload your audio to their servers to process it. For a recorded lecture, an interview, or a personal voice note, that never sat right with me — you're handing a copy of a private conversation to someone else. So ZoopZip does the opposite. Everything runs on your own machine using OpenAI's Whisper models (via whisper.cpp), accelerated on Apple Silicon (Metal) and NVIDIA (CUDA). Your audio and transcripts never leave your computer, and once a model is downloaded it works with no internet at all. Two people I really built it for: 🎓 Students — record the lecture, then turn it into clean, searchable text to revise from instead of scrubbing through an hour of audio. 🌍 Language learners — found a great video in your target language with no subtitles? Paste the link; ZoopZip fetches and transcribes it locally so you can read along and look up words. It's free — no account, no watermark, unlimited. There's a one-time (non-subscription) Pro upgrade for batch processing and the largest model, but the core is genuinely the whole app. Because there's no server, the privacy is the architecture — and it's also why the free tier can be unlimited. Mac & Windows: [zoopzip.com]. I'd love your honest feedback — especially on the privacy claims and where the free/Pro line should sit. I'll be here all day answering.

ZoopZip

Turn audio and video into text on your computer, offline.

Turn audio and video into text on your computer, offline.