Vibe — локальная расшифровка аудио и видео в текст

★ 7.1K

vibe is an open-source tool that transcribes audio and video to text entirely on your own machine. Reach for it when recordings must not go to the cloud: interviews under an NDA, medical appointments, legal consultations, internal meetings. The data never leaves the device. It accepts both audio and video, recognises speech with the Whisper, Nemotron 3.5 and Parakeet models, and returns the result in convenient forms — SRT and VTT subtitles, plain text, PDF, DOC or JSON for further processing. It works offline, so it also suits places without reliable internet. It goes one way: sound to text. This is a finished application with a UI rather than a library — a non-technical user can install and run it.