transcribe

To transcribe audio from a video file, the Skill automatically extracts the audio track and converts it to a standard WAV/16kHz mono format before sending it to the configured speech-to-text provider.

Yes, you can use OpenAI Whisper, Deepgram, or Google Gemini for speech-to-text. The active provider is selected in your Settings under Speech-to-Text.

To transcribe a local audio file, simply provide the local file path via the file_path parameter. The Skill handles common audio formats and routes them to your configured STT provider.

For robust transcription, this Skill enforces built-in timeouts and error handling to manage processing failures. This ensures reliable transcription of meetings, interviews, or podcasts.

Yes, the Skill supports chunking to process long audio and video files. This mechanism works alongside enforced timeouts and error handling to ensure robust transcription.

FFmpeg is needed to automatically extract audio from video files and convert it into a standard WAV/16kHz mono format, which is required by the speech-to-text provider for accurate transcription.