About
Lazy Captions is an iPhone app that automatically generates and burns subtitles into any video — entirely on-device, with no internet connection required. Using OpenAI's Whisper speech recognition model running locally, the app transcribes spoken audio and renders styled captions directly into the video file. The result is saved straight to the user's gallery. There's no cloud processing, no account, and no subscription — just a one-time purchase for unlimited use. An Android version is currently in development.
How It's Made
On-Device AI
Whisper AI runs fully on the device using Core ML. Four model sizes — Tiny to Turbo — let users trade off speed against accuracy depending on their needs.
Video Processing
Built with AVFoundation for audio extraction and video rendering. Captions are burned into the video frame-by-frame and exported in the original format.
Privacy First
No data ever leaves the device. No account required. The app works completely offline — including in airplane mode.
Gallery
Model Selection
Choose a Whisper model and highlight colour before captioning.
Processing Pipeline
Step-by-step progress: model load, audio extract, transcription, render, save.
Done
Confirmation screen after captions are burned in and saved to gallery.