Real-time multimodal chat with XENLOOK AI agents across text, voice, images, memory, and canvas workflows.