Since Figma Make became publicly available, I’ve noticed a significant drop in performance and overall intelligence. I’m not sure if it’s a coincidence, but I no longer see the “Claude Sonnet 4” label next to the prompt area. Everything I try now leads to poor results. Just yesterday, during the beta phase, the AI was much more capable and consistently delivered great outcomes. Today, I’m stuck for hours on simple tasks that were handled effortlessly before.
Anyone noticing the same?
Question
What happened to Figma Make?
Hi
I understand you're experiencing a performance issue in Figma Make file today, and you've also noticed the "Claude Sonnet 4" label is no longer visible.
Could you please try duplicating your Figma Make file to see if the performance improves?
If the issue persists, to allow our team to investigate this further, please fill out this quick form: https://figma.bot/submitreport.
In the form, please share the URL of the file where you're seeing this issue and grant edit access to support-share@figma.com. (Please rest assured that granting this access will not impact your billing.) This access will help us reproduce and investigate the problem effectively.
Thanks,
Yes lets not pretend that this isn’t intentional, as they’ve released to the wider Figma community and it no longer lists the “Claude Sonnet 4” label. Figma Make now struggles with the most basic tasks and more often than not decides to delete the entire application and create a smaller application purely focusing on the area prompted, despite being told not to do this.
The model used has definitely been switched. This, along with the ridiculous prompt limit, is going to force our entire organisation to jump ship.
it’s just a tactic to charge users to buy more credits. RIP , not using this. I rather switch to Loveable.
I have noticed that the queries do not get near the same quality result in make today. It also isn’t able to handle the depth of querying I was doing with it last week. This week the same query skews my entire app and changes things that I didn’t even ask to have changed.
It would be great if Figma could provide some level of explanation to the difference we’re noticing in the app since announcing that it is out of beta.
I’m also curious to understand how credits will work if the app gets it wrong and changes things we didn’t want changed with a query. Do we then have to use our credits to fix what was broken, even when we didn’t ask for it?
Hi there,
Thanks so much for taking the time to share your experience and feedback — I’ll pass it along to the team.
We’re currently offering multiple credits so you can continue creating and refining your results. We’re also planning to offer the ability to purchase additional AI credits later this year. Until then, we’re not strictly enforcing limits for paid Full seats. You can check out the How AI credits work Help Center article for more details.
We’ve also published some best practices for Make. Please try applying a few of those tips to get better outputs with fewer prompts.
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.