One key to voice, video, and agents
Bearer auth and a streaming endpoint return expressive audio while the words are still being spoken. Render talking characters from a script, open live agent sessions with a single call, and run dubbing on the same pipeline — the surface area stays small enough to hold in your head.
