Every AI-assistant feature in the platform just got an upgrade, and one long-standing Motion AI limitation is gone.
The model upgrade needs nothing from anyone — no toggle, no migration. Auto-generate formats, the Attention Heatmap, and every other AI-assistant feature simply produce better output than they did last week, at the same cost to us.
The SVG change removes a manual step designers have lived with since Motion AI launched. Previously, when a prompt needed an image layer, Motion AI dropped in a rectangle as a stand-in — the designer then had to delete it, upload the real image, reposition it, and match the settings by hand. Now Motion AI builds the SVG directly, ready to use or swap out. One prompt, one fewer round trip back to the file.
The Figma → Studio pipeline has had its biggest overhaul yet, with a new UI and a longer list of what it can actually handle.
Complex Figma files used to be the edge case that broke the import. This release closes most of that gap: more layer types come through intact, errors surface with an actual explanation instead of a dead end, and updating a template no longer means starting over from scratch.
The net effect is fewer manual fixes between "designed in Figma" and "live in Studio." A full changelog is available in Notion for anyone who wants the detail.
Motion AI now ships with a library of ready-made prompts, built directly into the platform. A blank prompt field only helps the people who already know what to ask for — the library puts the possibilities in front of everyone else instead of leaving them to guess. It's built so new prompts can be added without involving development, so if a prompt would help customers, it can be live the same day.