Not sure which mode or model fits your idea? Describe it. The assistant translates plain language into the right generation and drops the result onto your canvas.
Describe it in plain language
"A drone shot flying over neon Tokyo at night" is enough. No mode-picking, no parameter hunting — just say what you want to see.
It picks the model for you
The assistant chooses the mode and model that suit the request, writes the prompt, and generates — then hands you a shot you can drop straight onto the timeline.
Refine on the canvas
Every result is a node you can branch, re-prompt, or feed into the next model. The assistant gets you started; the canvas lets you take it further.