Speech-to-Text Implementation

Sypha offers experimental support for auditory input, enabling you to dictate your technical requests rather than typing them manually.

Environment Preparation

1. FFmpeg Engine: Essential for processing audio capture.
   • macOS: brew install ffmpeg
   • Linux: sudo apt install ffmpeg
   • Windows: Locate binaries at ffmpeg.org and ensure the executable is in your system PATH.
2. AI Transcription Provider: This feature is powered by OpenAI's Whisper engine. You must verify that your OpenAI credentials are active and configured within the Sypha profile settings.

Enabling Voice Input

1. Activate the Speech to Text prototype in Settings -> Experimental Features.
2. Locate the microphone widget now available in the chat input bar.
3. Record your message by clicking the icon; select it once more to finalize the capture.
4. Your dictated request will be instantly processed and populated in the input field.

Common Resolutions

• Interface Widget Unavailable: confirm the experimental toggle is active and verify that FFmpeg is correctly installed and accessible via the command line.
• Service Interruptions: Ensure your OpenAI account has sufficient credits and your network connection is reliable for data transmission.