Use cases

Coordinate image, audio, and video agents to check for one object

Google's multimodal tutorial uses a LangGraph orchestrator to send files to specialist Gemini agents, gather their modality-specific findings, and synthesize whether a requested object is present.