Last released Oct 18, 2026
Open, model-free video context layer for LLM agents: index once, query cheaply with any model.