Last released Nov 22, 2026
Turn short videos into structured, timestamped text for LLMs: speech, on-screen text, and visual context.