Skip to main content

Hi Figma Community and Team,

I'm a developer (with limited UI skills) who's been absolutely loving Figma Make – it's a game-changer for me!

I'm thinking about upgrading to a Full Seat subscription mainly for this feature, but I'm trying to wrap my head around the AI credit system. The docs mention that *"about 50-70 Make prompts may consume 1 AI credit."*

I'm struggling to visualize what counts as a "prompt." For example, if I ask for something reasonably complex like:

"Generate a homepage for an e-commerce site. It needs to have both light and dark mode themes, a hero carousel banner, and display products in a card layout."

Does anyone from the team or community have a rough idea of how many credits a single, multi-faceted prompt like that might consume? Is it closer to 1 credit, or would it be more because it's packing several instructions?

My other big question is about the Full Seat subscription. I saw this note on the website:

"Until then, we won’t be strictly enforcing credit limits for Full seats, but limits will still apply for other seat types."

Does this effectively mean Full Seats have unlimited AI credits for now? And if we do somehow hit a limit, does access to Figma Make just get cut off?

Thanks in advance for any clarity you can provide !

Correction to my original post: My apologies, I misstated the credit amount. The official allocation is 3,000 credits per month, and the "50-70 prompts" is an estimate of what that total monthly credit pool might translate into, not a "per credit" cost. My core question about complex prompt consumption and the "unenforced limits" for Full Seats still stands. Thanks!

Hi ​@AXiu920 ,


I understand you're looking for more clarity around prompts and AI credits in Figma Make

 

Regarding your question about how many AI credits a complex prompt might consume.

I looked into it, but unfortunately, we don’t have any additional information to share beyond what’s currently listed in the table in How AI credits work

If you're concerned about credit usage, you might also find this blog post helpful: 8 ways to build with Figma Make 

 

As for Full seats: we're planning to offer the ability to purchase additional AI credits later this year. In the meantime, we’re not strictly enforcing limits on Full seats.

 

Thanks,

Hi ​@AXiu920 ,


I understand you're looking for more clarity around prompts and AI credits in Figma Make

 

Regarding your question about how many AI credits a complex prompt might consume.

I looked into it, but unfortunately, we don’t have any additional information to share beyond what’s currently listed in the table in How AI credits work

If you're concerned about credit usage, you might also find this blog post helpful: 8 ways to build with Figma Make 

 

As for Full seats: we're planning to offer the ability to purchase additional AI credits later this year. In the meantime, we’re not strictly enforcing limits on Full seats.

 

Thanks,

​Hi ​@Junko3 ,

Thank you for your reply; it has cleared up my confusion.

However, while using Figma Make today, a prompt appeared saying: "We won't start enforcing AI credit limits in Figma Make until later this year, but you've reached your future credit limit."

In your previous response, you mentioned, "In the meantime, we’re not strictly enforcing limits on Full seats." Does that mean I can still use the AI features?

Looking forward to your reply.

Thanks!

Hi Ax,

Figma Make is undeniably a powerful and game-changing tool. I'm genuinely impressed by what it can do. That said, I’ve found the current credit system to be somewhat impractical when working with more complex systems.

I got access last Thursday and have already burned through my prompt credits with just a full seat. I had great success building several smaller prototypes, but when I tried a more complex, EPIC-level prompt involving Supabase API connections, things got tricky.

It’s hard to tell whether the issue was the complexity of that single prompt or the cumulative total of prompts, as described in the documentation. Either way, both small and large projects tend to generate numerous “Fixes,” which consume additional prompts—often without resolving the issue effectively. This iterative fixing process is what ultimately drained my credits over time.

Personally, I don’t think “Fixes” should count against the credit limit. I probably used up half my credits just trying to correct issues. For example:

  • I had to manually resolve an API key issue outside of Figma Make after burning six prompts trying to make a small change work.
  • Another case involved a toggle that wouldn’t update correctly—it took five prompt attempts before it finally worked.

These kinds of troubleshooting loops make it difficult to experiment meaningfully with advanced use cases. I’d love to hear others thoughts or if there are any best practices for managing credits more efficiently in these scenarios.

Best,
JP

Hi Ax,

Figma Make is undeniably a powerful and game-changing tool. I'm genuinely impressed by what it can do. That said, I’ve found the current credit system to be somewhat impractical when working with more complex systems.

I got access last Thursday and have already burned through my prompt credits with just a full seat. I had great success building several smaller prototypes, but when I tried a more complex, EPIC-level prompt involving Supabase API connections, things got tricky.

It’s hard to tell whether the issue was the complexity of that single prompt or the cumulative total of prompts, as described in the documentation. Either way, both small and large projects tend to generate numerous “Fixes,” which consume additional prompts—often without resolving the issue effectively. This iterative fixing process is what ultimately drained my credits over time.

Personally, I don’t think “Fixes” should count against the credit limit. I probably used up half my credits just trying to correct issues. For example:

  • I had to manually resolve an API key issue outside of Figma Make after burning six prompts trying to make a small change work.
  • Another case involved a toggle that wouldn’t update correctly—it took five prompt attempts before it finally worked.

These kinds of troubleshooting loops make it difficult to experiment meaningfully with advanced use cases. I’d love to hear others thoughts or if there are any best practices for managing credits more efficiently in these scenarios.

Best,
JP

​Hi ​@John_Petersen ,

Yes, I agree with your point.

For me, the purpose of using Figma Make is solely for UI design, since my project is not developed with React—it uses Vue/Nuxt.

When Figma Make generates interfaces for me, it sometimes produces warnings or errors. Since I'm not familiar with React, fixing them on my own is very difficult, and using Figma Make to fix them consumes AI credits, which is clearly unreasonable!

Terms & ConditionsCookie settingsAccessibility statement