I’ve tried Figma Make and have struggled to get any meaningful results from it in the areas of my current process where it has the biggest opportunity.

Other AI prototyping/LLM tools already bridge the gap for rapid, rough prototyping, exploring ideas, trying multiple approaches quickly, etc. All things that happen early on in the design process. What they can’t do, however, is closely match my designs (whether it uses my system components or not) visually (and quickly without forcing me to prompt it dozens of times about surface level issues). I assumed this is what Figma Make would capitalize on, but it falls very short.

Figma Make is the most well positioned to take something I’ve already designed in high fidelity and make a robust, interactive prototype of it. It doesn’t even need to be working code that I can hand to a developer, that’s a different problem. But a prototype that clearly shows detailed (oftentimes complex) functionality that I can use to demo for devs, usability test with, and test for edge cases with would be a game changer. Today, it will basically revert my designs back to the stage I already had them with early prototyping exploration which I can do with any other tool on the market. I want Figma Make to get me from 80-99% when it comes to fidelity.