Lugs

Lugs generates live captions and transcriptions of computer audio and microphone input entirely on your device, with no internet connection required. Built by a hearing impaired team, it targets people who need accurate, private captions during real conversations rather than cloud based transcription suites.

Lugs is a desktop captioning and transcription tool that puts subtitles on the audio your computer is already playing, plus whatever comes in through a plugged in microphone. It runs the speech recognition model locally, so the audio from a call, a video, a lecture or an in person conversation is processed on the machine rather than streamed to a remote server. The headline claim is that it works with no internet connection at all, which makes it unusual among captioning products that generally depend on cloud inference. The stated audience is broad but the design clearly favours people who need captions to follow a conversation: users with hearing loss, students in noisy rooms, remote workers who miss words on a call, and anyone who wants a written record of spoken audio without uploading it anywhere. The team behind it describes itself as hard of hearing and says the tool is built around its own daily use, which explains the emphasis on accuracy in ordinary dialogue rather than on studio grade transcription features. In practice the workflow is simple. You download the app for your operating system, plug in a microphone if you want to capture speech in the room, and Lugs starts producing live captions over the running conversation. The product claims it follows the context of a dialogue rather than treating each utterance in isolation, which is how it aims to keep names and half finished sentences readable. Access is described as early access, and the site promises lifetime updates with the purchase. Positioning matters here. Cloud transcription services typically offer editing suites, speaker labels and export pipelines but require you to send audio off your device. Lugs trades that breadth for privacy and offline reliability, which suits live listening far more than post production work. Anyone expecting a full transcript editor or meeting summaries should look elsewhere; anyone who wants captions to appear while the audio is happening, without a network, is the target user. Pricing and platform availability beyond macOS are not spelled out on the homepage, so confirm details before committing.