audio-transcription-workflow

Use the transcription workflow to process single files, folders, or recursive glob patterns, converting diverse audio formats into structured markdown transcripts with metadata headers for easy tracking.

Yes, local Whisper handles primary transcription, with an automatic OpenAI API fallback and OpenRouter-based post-transcription text refinement available for cleanup and formatting.

Yes, you need local Whisper access for primary transcription, OpenAI API keys for fallback, and OpenRouter API keys for post-transcription text refinement and cleanup of voice notes.

Process interview recordings through the multi-stage pipeline: local Whisper transcribes, OpenAI provides fallback, and OpenRouter cleans up the text into structured markdown files with metadata headers for research documentation.

Yes, the workflow supports recursive folder processing and glob patterns, enabling efficient batch transcription of large audio archives into structured markdown transcripts with metadata headers.

AI cleanup via OpenRouter refines raw speech-to-text output into structured markdown, ensuring voice notes and interview transcripts are immediately readable and ready for further documentation or AI analysis.