AirCaption

What it does

AirCaption turns audio and video into captions and transcripts on your own machine. It uses NVIDIA, Cohere or Whisper models and works entirely offline by default. You can edit text and timing, subtitle video in more than 80 languages, and export SRT, VTT, plain text or burned-in captions. It runs on Mac and Windows.

AirCaption is a desktop app for Mac and Windows that transcribes audio and video locally. The homepage says everything runs on your machine out of the box and nothing is uploaded unless you choose it. You can generate captions, then fix words and timing with hotkeys, find and replace, and a waveform view. Existing caption files can be imported and edited, and whole folders can be dropped into a queue. Optional cloud translation sends only the caption text you select. The pricing page lists a free Essentials plan with unlimited AI transcription and 10 minutes of cloud translation per month, and a Pro plan at $9.99 per month that adds medium and large AI models and 30 hours of cloud translation. The homepage names video editors, podcasters, language learners, legal professionals, marketers, researchers, event organizers, online course creators and journalists as users. No public API or open source license is stated on either page.