[Config 2025] Figma Make is here — let’s hear what you think! 🎨

I was VERY Impressed with the capabilities of figma make to start but after playing with it for a couple weeks i feel like i’ve hit a dead end. Not sure how this saved me any time or improved my workflow. 

 It was very buggy even when I gave specific components to use or specific screenshots and told to replicate exactly it would often do its own thing / version of it.  So now I have a really fun interactive prototype but I still have to fine tune to use for our design system and to create a hand off file for developers.

To make minor tweaks (even just removing an element or changing the wording on something) required so much time for the code to re-write itself. Essentially was always taking a gamble on whether it was going to do what I said, or take what I said to do, and recreate things that I didn’t ask for.  I constantly had to give specific instructions to “not make any other changes etc” and hope for the best. 

Exciting and understand its still in BETA so I hope these key improvements happen to make this useable for designing products at scale and to solve real problems for real key users (designers). 

Does Figma Make have an API?

Wow is this product concept disappointing.

• Claude can’t remember what it did 10 minutes ago and can’t reference prior code
• Creates more problems that it solves for after every prompt
• Rewrites entire chunks of code that was working and is now not working (code sections you didn’t ask it to rewrite, nor were related to the prompt’s ask)
• Spends more time analyzing your prompt and recompiling that you do actually creating anything with it - wasted time for UX and UI
• Doesn’t understand contextual conversation, you have to be very literal of everything you say. I asked it to update an input and instead it added a second input with the code fix.

For example:

• “Create a modal with these specs” prompt
• one hour later you reopen the modal having not prompted anything to do with it and the content is completely different. You never asked Claude to update the modal. Now there is a new form title, new form elements, completely different layout inside. what was functioning before is now not functioning.

As someone who develops and designs - this product isn’t worth much. I know it’s beta, but damn. If your target audience is PM’s they won’t use it because there are way better options out there.

Imagine I said “the car is low on gas” and AI decided to remove the tires because to it logically, this is how you prevent any more gas from being used. That’s how this product has reacted to my prompts.

Also, the compiling is slow, you can’t easily edit anything, you can’t delete a file (and the AI isn’t permissioned to either!)

This product is about 2 years too earlier of a release to be useful. I thought this was supposed to be a replacement for how absolutely underdeveloped and broken Figma prototyping is. I was using it so to reflect real code and inputs, and not deal with all the bugs of the main prototyping tool when user testing.

I know you probably put a lot of love and effort into this...

But PLEASE Figma, No more new products. Please focus on your core product!

I wish I could edit in Figma what I create in Figmamake… i feel like I stopped being a designer and must become a prompter. I feel envy that my developer colleague is having a great time. Because he now can do a prototype in a day, which would take him a week before. I am getting less tasks because now they can be created in Figma Make by developpers. I genuinely hate the feature. 

But I am thankful for Figma Sites, they are like Framer but in Figma. Breakpoints in fonts styles are a game changer. 

I personally think Figma Make is great since I have hired no less than 5 designers to create my layouts and after a month, I still don’t have my UI done.  I don’t really understand designers who say they will get something done and then don’t look work on my project for days (some never even started working) or work 1 hour a day here and there. I tried inexpensive and expensive designers and had the same results: they left me waiting and providing the exact same feedback over and over (wasting my time).

When I discovered Make, I realized I could do a create the screens myself and then, I was super excited I could create actual functional interfaces!  Super cool. I’m becoming a prompter for sure (I’m not a good developer either so this is perfect for people like me with little skills).

After using Make for a few hours: I fired my designer (who wasn’t doing anything anyway) and now I’m doing it myself.  I have developers waiting to see how this turns out (as am I).

ONE NOTE: The response has gotten a lot slower. I’m on Version 149.  The AI makes the same mistakes over and over, which causes delays and prompts to stop making the same mistake… but it keeps making the same mistakes regardless of my prompts.  Yes - I added a guideline to tell AI to avoid doing X but it still does X.  But this is still moving forward faster than my former designers were.

Thanks Figma.

We need to be able to edit make designs as a figma file. I feel like that would be absolutely perfect collaboration client/designer. Until then, it’s a nice toy to play with but I will have to replicate it manually if I want to modify specific parts without re-prompting over and over 😥

Currently, I found it to be a nice prototyping agent.  However, it stops there.  How can you take that project and export it to add code for a website or app?  What’s the utility if you can’t do so?  

right now, it’s a cool fun little toy, but in terms of utility in a development stack, it’s minimal to non in terms of usage.  I’m starting to feel sorry I paid for it.  

You’re absolutely right.  I paid for it and right now I’m starting to regret it.  Until you can edit your designs or export it according, it’s virtually useless.  

Hello Figma Team!

One thing I would like as an admin of Figma for our organization is the ability to turn Figma Make on or off for individual users. As it stands, it is either accessible for all full users, or disabled for all full users. We currently have is disabled because we don’t want users bypassing our UX Teams by using AI functions instead. But having the flexibility of giving this to our AI Team, or specific users would be great so we could benefit from the feature.

Thanks!