WhisperWhy – Local STT WhisperFlow Alternative (github.com)

WhisperWhy has launched as a local-first speech-to-text (STT) tool for macOS, offering users a secure way to dictate text from anywhere in the operating system without their audio or transcripts leaving the device. By simply pressing a configurable hotkey, users can capture their speech, which is processed on-device using whisper.cpp for transcription and cleaned up by a local LLM (by default, Ollama). The complete application does not display the transcription but pastes the finalized text directly into the active text field, enhancing privacy and user experience. This tool is significant for the AI/ML community as it emphasizes a local processing approach in an era where cloud-based services dominate STT technologies. The use of Metal-accelerated whisper.cpp for STT and the flexibility to integrate any OpenAI-compatible LLM for text cleanup highlight the potential for high-performance, local AI applications. Furthermore, with built-in features such as a customizable hotkey, an animated user interface, and an array of settings for developers, WhisperWhy exemplifies the growing trend towards privacy-centric AI solutions that harness the power of local computation.