Voice dictation is here
Talk to your codebase. Anibuild now supports voice input for hands-free coding sessions.
Starting with Anibuild v0.2.100, prompts can be dictated instead of typed. The feature uses native speech-recognition APIs and keeps audio on the local machine.
01 // How it works
Press Ctrl+V in the prompt to toggle voice mode. Anibuild listens through the default microphone, transcribes in real time, and submits the transcription when you pause.
“Move all the database queries in the user controller into a separate repository class and update the imports.”
02 // Supported platforms
- macOS — Uses the built-in Speech Recognition framework.
- Linux — Uses PipeWire/PulseAudio with Whisper.cpp for local transcription.
Windows support is planned for a future release.
03 // Configuration
Customize voice input in the config file:
voice:
enabled: true
language: en
silence_timeout: 1500
model: whisper-small