Audio Transcription Use When - The input is audio and the main job is speech-to-text, subtitle-ready timing, rough speech search, multilingual transcription, or durable transcript artifacts. - Use for video inputs and for semantic video understanding. If a timed transcript already exists and only subtitle output is requested, skip transcription and read the local subtitle conversion reference. Do Not Use When - The task needs new creative generation or visual analysis rather than speech or subtitles. - Required inputs are missing and guessing would change the result. Execution Boundary - Host…