CREATE INTERFACES FROM A PROMPT: FIGMA GETS GEMINI 3 PRO
Google's new Gemini 3 model is now integrated into Figma Make, where it can generate layouts, components, design systems, and even HTML/CSS from text prompts or reference images. It speeds up prototyping with animations and states, and while it still needs refinement for complex work, it significantly lowers the barrier for beginners and boosts efficiency for pros.