Typewhisper: Local speech-to-text on-device AI, private, optional cloud (github.com)

TypeWhisper has launched a new version of its local speech-to-text application for macOS, allowing users to transcribe audio using advanced on-device AI models while maintaining data privacy. This release includes a suite of features like system-wide dictation, file transcription, and customizable workflows that streamline the transcription process. Users can choose between processing their data locally on the Mac or utilizing cloud APIs, such as Groq and OpenAI, for faster results. Notably, the app also facilitates real-time transcription previews and supports batch processing of audio and video files. This development is significant for the AI/ML community as it enhances accessibility to speech recognition technology, promoting both user privacy and versatility in data handling. With support for ten different engines, including Whisper and Qwen3, TypeWhisper caters to a wide range of languages and processing speeds. The app’s robust workflow capabilities enable users to implement automation tailored to specific tasks, fostering productivity in professional settings. Overall, TypeWhisper represents a meaningful step in making powerful speech recognition tools more user-friendly and customizable while ensuring data confidentiality.