Speech-to-Text Linux
Enables local speech-to-text functionality for Claude Code on Linux, integrating voice input directly into the AI's conversation stream.
About
This tool provides a local Speech-to-Text (STT) server specifically for Linux, designed to integrate seamlessly with Claude Code via the MCP protocol. It allows users to control Claude Code with their voice through a push-to-talk mechanism, using the Right Ctrl key to activate recording. Speech is transcribed using the Whisper tiny model by default, and the resulting text is directly injected into Claude's input stream within a Tmux session. The server runs in a Docker container, accessing host keyboard and audio devices to facilitate its Linux-exclusive functionality.
Key Features
- Push-to-talk (PTT) voice input activated by the Right Ctrl key
- On-device speech-to-text transcription using the Whisper tiny model
- Direct injection of transcribed text into Claude Code's input via Tmux
- Dockerized server for isolated and consistent operation
- Modular architecture for flexible component management
- 1 GitHub stars
Use Cases
- Hands-free coding or command input within a Linux environment
- Interacting with Claude Code using voice commands and dictation
- Improving workflow efficiency for developers by reducing manual typing
Sponsored