Agents can make video now
Connect fal or xAI, put a video model on a task in the Rack, and ask an agent for a clip. It lands in the reply, ready to play.
The Rack has a new Video generation category with two tasks: text to video and image to video. Connect fal or xAI in AI Gateway, then put one of their video models on the task you want.
From there you just talk to the agent. It can draw the stills with your image model, film each one with your video model, and hand back the clips in the conversation. Each clip is saved in your project and plays right in the reply.
Providers charge per second, so a clip is five seconds unless you ask for another length. Stop works while a clip is filming. The Groups section at the bottom of the Rack is gone, because each task now simply has its own model.