Openasr

OpenASR is a local first speech to text application for macOS and Windows, with Linux listed as next. It turns speech into text on the user's own computer, with no account, no cloud processing and no audio upload. Models, audio and transcripts all stay on the device, and the app keeps working without a network connection once a model has been downloaded. The desktop app offers three ways to use it. Files mode transcribes meetings, interviews, lectures and voice memos, with inline editing and export as plain text, structured data or timestamped subtitles. Dictation types words wherever the cursor is, using a hold to talk shortcut and adding punctuation automatically. Live captions show a floating overlay that follows online lectures, calls and podcasts in real time. Other features include speaker separation for meetings and interviews, timestamps kept beside each line, a choice of lighter or larger models, and a history of recent files and results. The core engine is open source, and the same engine is available through a command line interface for scripting from a terminal. The desktop app is the simpler path for people who want no setup. The page is also available with a Chinese language version. The site shows no paid plans. Structured data on the page lists a price of zero, and no subscription or account is required to use the app.