OSTT – Visual terminal based speech-to-text for Omarchy/Any Linux/Mac (github.com)

OSTT, a new terminal-based audio recording and speech-to-text tool, has been announced for Linux and macOS. Built with Rust, OSTT enables users to record audio with real-time waveform visualization and transcribe it automatically using various AI transcription models. Key features include audio clipping detection, adjustable reference levels, the ability to manage transcription history, and a customizable interface that allows it to run as a floating popup window in any app. Users can choose from multiple AI providers such as GPT-4 and Whisper models by configuring their own API keys. This development is significant for the AI/ML community as it offers a lightweight, efficient solution for audio transcription that integrates easily into existing workflows. The versatility of using different AI providers broadens its applicability for various transcription needs, while the capability to manage vocabulary through keyword support enhances accuracy, particularly for technical jargon. With its cross-platform compatibility and minimal dependencies, OSTT positions itself as a valuable tool for developers and users looking for a straightforward, powerful audio transcription solution.