New Member

Figma Make Ai Credit Limits Not Feasible

Forum|Forum|6 months ago
March 10, 2026
187 replies
9321 views

I’d like to share feedback regarding the implementation of AI credit limits in Figmake. As a designer managing multiple complex projects, I believe credit restrictions would greatly undermine Figmake’s usefulness in real-world design and prototyping workflows.

1. Most AI usage is for fixing bad or incomplete output, not luxury use.
In practice, a large percentage of AI prompts in Figmake are iterative—used to correct layout issues, rebuild misinterpreted elements, or align behaviors with visual intent. Each refinement quickly consumes credits even though the output quality depends on that iteration. Penalizing iteration limits the model’s ability to produce accurate prototypes, especially when the AI output is imperfect through no fault of the user, making credit usage really unfair and frustrating.

2. Predictability becomes impossible, discouraging adoption.
With a credit cap that no one knows how to predict and calculate ahead of time, I can’t plan how far my Figmake usage will take me in a given project. Credits could run out mid-way, forcing me to manually rebuild AI-generated prototypes and creating major inefficiencies. For example, if I spend a week using Figmake for an interactive prototype and then hit a credit limit right before stakeholder review, I’d end up having to redo everything by hand including component mapping and manual prototyping—rendering the AI tool more of a liability than an acceleration.

3. I actively use Figmake for time-intensive, pixel-perfect projects.
I currently have five ongoing projects that rely on Figmake’s automation to achieve understandable prototypes across varying layouts, dynamic conditions, and complex component interactions. These require functional, clickable prototypes that would otherwise take hundreds of manual steps. To give you a sense of credit usage to achieve a prototype, each project reached hundreds of versions within a week of creation. AI limits directly restrict my ability to complete these on schedule or maintain parity between design and prototypes.

4. Prototypes must stay updated and accurate.
Projects evolve rapidly across teams. If AI updates become throttled because of credit exhaustion, Figmake prototypes drift from reality—outdated screens, missing conditional flows, untested interactions, and inconsistent visuals make it impossible to keep stakeholders aligned. Designers would revert to traditional Figma, eliminating Figmake’s intended value.

5. Figmake bridges a huge gap with non-designer stakeholders.
Stakeholders don’t want to learn how to use Figma’s preview tools—they struggle with finding files, switching flows, adjusting zoom, or enabling hotspots. Figmake solves this by letting me send a simple link everyone can understand without training. Credit limits would prevent consistent sharing of working AI-powered prototypes, creating friction in alignment and communication with non-design teams.

6. Credit-based systems penalize professional users.
Professional users already pay for Figma Organization plans come with the expectation of scalability and reliability. AI credit limits undermine that—power users end up penalized, while lighter users are unaffected. Enterprise workflows require consistency, not unpredictable usage caps. Requiring designers to repeatedly request additional credits from admins adds unnecessary friction, forcing us to justify needs and placing stress on teams with little control.

7. Creativity and exploration get stifled.

Part of Figmake’s value lies in exploration—testing concepts, states, and flows in seconds. Predictive anxiety over “credit drain” limits experimentation. Designers will stick to safe, minimal use rather than exploring variations that lead to innovation.

8. AI credit management adds unnecessary overhead.
Tracking remaining credits adds cognitive and administrative burden. Instead of focusing on design, we’d have to monitor dashboards, request top-ups, or pause work mid-flow. This complexity erases the simplicity and joy that make Figma tools beloved.

Recommendation:
Instead of credit limits, please consider providing unlimited AI usage under paid plans with limits only for automation abuse.
- Allowing local refinement of previous AI generations without repeated credit consumption.

Pinned Reply By ksn

Hey @baewadalmaster @DonDesigns @JaredSF - we have been consolidating similar threads so conversations and updates around the same topic stay in one place. In this case, we’re merging related discussions into this thread. This makes it easier for people to quickly find answers to their questions.

When topics are merged, some threads may be closed or redirected, which can make it seem like things have disappeared (especially from older links).

If one of your posts is merged into another topic, you can always find it through your profile and see where it now lives.

If you have any other questions, let us know.

Show previous replies

Maxter

New Participant

I’ve burnt 34000 credits in 4 days. Having 3000 credits/month is ridiculously low. That would have been $900. With this rate, in a month I would use almost 300.000 credits which is 9k USD. I am leaving Figma on March 18th.

AaronUX

New Member

I’m writing as a paying Professional-plan user who has been using Figma Make heavily to build and maintain a portfolio website. I’ve found Make genuinely useful, fast, and creatively enabling, which is exactly why the current AI credit model is so concerning to me.

My main issue is not just that limits exist. It is that the current system makes usage feel unpredictable, punitive, and poorly aligned with how a conversational creative tool actually works.

In practice, Figma Make is not a one-shot utility. It is iterative by nature. It often takes multiple rounds to clarify intent, correct misunderstandings, fix bugs, refine interactions, or recover from mistakes in the output. That is normal use, not abuse. When every clarification, correction, retry, and follow-up consumes credits, the tool starts to feel less like part of my subscription and more like a metered gamble.

That creates a few major problems for me:

It discourages natural creative exploration, which is exactly where a tool like Make is most valuable.
It penalizes users for model misinterpretation, failed attempts, and normal iteration.
It makes cost difficult to predict in advance, which creates anxiety around using the tool freely or conversationally.
It even discourages basic politeness or rapport in the interface, because every extra message can feel like a waste of limited credits.
It is especially frustrating in a design workflow, where iteration is not optional - it is the work.

I recently used almost all of my AI credits in about one morning of legitimate work updating my personal site. That is not because I was using the tool irresponsibly. It is because I was working with it the way it seems designed to be used: interactively, creatively, and through refinement.

I understand that AI features have real costs. But as implemented, this system feels too opaque and too restrictive for a tool that is still imperfect and often requires multiple attempts to get to a usable result. Charging per-use-like costs for a workflow that inherently depends on retries, interpretation, and debugging makes the tool feel stressful rather than empowering.

I also want to specifically suggest a flat monthly pricing option for individual Professional users. The rest of the Professional plan is already structured as a straightforward subscription, and that model fits how creative tools are normally used. Right now, the absence of a predictable flat-fee option for solo users feels like a major gap.

A flat monthly AI option for individual users would make far more sense than forcing people into an unpredictable credit system for normal use. I understand charging more for heavy enterprise usage, large teams, or unusually high-volume workflows. But for independent professionals, job seekers, freelancers, and contractors, the current structure makes Make feel financially risky to rely on, even while already paying for a Professional seat.

What makes this especially frustrating is that other subscription products manage to absorb variation in individual usage without turning normal engagement into a stressful meter. Some people use a service lightly, others use it heavily, and the flat monthly price is part of what makes the product feel usable and trustworthy. For a creative assistant built into a design platform, that kind of predictability matters.

I’d strongly encourage Figma to consider changes such as:

a flat monthly AI option for individual Professional users
more generous included credits for active Make users
clearer per-action cost visibility before a request is sent
lower-cost or free handling for retries, corrections, and failed outputs
some form of session-based or workflow-based pricing rather than charging every conversational turn so aggressively
better protection against burning credits on normal back-and-forth required to make the tool usable

I’m sharing this because I like the product and want it to succeed. Make has been one of the most useful parts of Figma for me recently. But the current credit model makes me hesitant to rely on it, and that is a bad outcome for both users and the product.

DonDesigns

New Member

I’m a professional designer and my company is spending time this week exploring alternatives to Figma Make since we can easily hit 30k in 3-4 work days. Did Figma think companies would be willing to pay that level of a premium when other options exist for so much less?

The bulk of those credits are spent correcting issues that Figma Make itself introduces, and there are some days where their AI clearly just doesn’t work at all, so even more credits get used up. I’ve also had it use a crazy amount of tokens just to resize something very small, by a few pixels.

jeffk

New Member

In 1 hour, all of my tokens were used up for the entire month. This token usage model will make Figma Make unusable. The minimum that I see is 150 tokens per request. Is this correct? It should be more like 1 to 3 tokens per request.

A

Anonymous

Im out, moved all front-ends to vercel, kept the supabase backends, GitHub holds the repos, and im using opencode to do the prompts, its quick, accurate and i have made 10 times the progress in 24 hours then what i did with Figma, there are alternatives, the cost point is very different as well as the ability to use any models i want without restriction. Its actually made the deploys much more efficient with extended error checking.

Alente

New Member

You've got to start with the customer experience and work backwards to the technology. We are giving figma valuable inputs on the user side. Sad to say they are not hearing it. I am trying claude code today, make was such a fine tool but I think its just in the wrong company i guess.

martijnmartens

New Member

Got a mail from Figma about the AI credits limits. Obviously, smart move, knew this was coming.

However, I have a couple of things here that rubs me the wrong way. Hope this reaches the higher-ups:

Additional credits are too expensive compared to the pricing/license model.

Credits are now per full seat. I have 11 full seats, only 2 or 3 of them are using the credits. I does not make sense to buy additional credits when I have about 8 full seats with 3000+ ai credits that are not being used.
1. Let us just share the credits we have across seats, instead of us sharing login details to maximize the credits.
2. If this is not the case, its better to buy just buy more full seats instead of credits. License for a full seat is 16 pm with 3000 credits, 5000 credits are 120 p/m? So additional credits are 4.5x more expensive then just buying full seats... This does not make any sense.
In the full seat, I'm paying for a lot I don't use, but then have to pay additional for Ai credits that I do need. Can there please be just a design license? The only thing we use is Figma design and Figma make. We would like to only purchase that, we don't need the rest. Its like buying a house with 10 rooms but I'm only using 2. But then you make these 2 rooms more expensive, but I cannot get rid of the 8 rooms I dont use.

Update: after using figma make for 1.5 hours. All the credits on my full seat are gone. Product is not ready yet. This makes Figma make completely useless honestly.

Fix this…...

Isabelle Peng

New Member

Im out, moved all front-ends to vercel, kept the supabase backends, GitHub holds the repos, and im using opencode to do the prompts, its quick, accurate and i have made 10 times the progress in 24 hours then what i did with Figma, there are alternatives, the cost point is very different as well as the ability to use any models i want without restriction. Its actually made the deploys much more efficient with extended error checking.

@kitwalker14 I was planning the same.. how do you manage the image bedding issue from Figma? the links in repo apparently will never work if directing to figma still..

Isabelle Peng

New Member

This credits hoax is going to driving the users away, I’m pro annual sub, if this is not workable, I will start solely on opencode myself, and downgrade my sub in figma.
Totally disappointed!
They are somehow losing money I get it, and want to monetize as much as possible, but this is going to backfire.
💀💩

Isabelle Peng

New Member

I’m writing as a paying Professional-plan user who has been using Figma Make heavily to build and maintain a portfolio website. I’ve found Make genuinely useful, fast, and creatively enabling, which is exactly why the current AI credit model is so concerning to me.

My main issue is not just that limits exist. It is that the current system makes usage feel unpredictable, punitive, and poorly aligned with how a conversational creative tool actually works.

In practice, Figma Make is not a one-shot utility. It is iterative by nature. It often takes multiple rounds to clarify intent, correct misunderstandings, fix bugs, refine interactions, or recover from mistakes in the output. That is normal use, not abuse. When every clarification, correction, retry, and follow-up consumes credits, the tool starts to feel less like part of my subscription and more like a metered gamble.

That creates a few major problems for me:

It discourages natural creative exploration, which is exactly where a tool like Make is most valuable.
It penalizes users for model misinterpretation, failed attempts, and normal iteration.
It makes cost difficult to predict in advance, which creates anxiety around using the tool freely or conversationally.
It even discourages basic politeness or rapport in the interface, because every extra message can feel like a waste of limited credits.
It is especially frustrating in a design workflow, where iteration is not optional - it is the work.

I recently used almost all of my AI credits in about one morning of legitimate work updating my personal site. That is not because I was using the tool irresponsibly. It is because I was working with it the way it seems designed to be used: interactively, creatively, and through refinement.

I understand that AI features have real costs. But as implemented, this system feels too opaque and too restrictive for a tool that is still imperfect and often requires multiple attempts to get to a usable result. Charging per-use-like costs for a workflow that inherently depends on retries, interpretation, and debugging makes the tool feel stressful rather than empowering.

I also want to specifically suggest a flat monthly pricing option for individual Professional users. The rest of the Professional plan is already structured as a straightforward subscription, and that model fits how creative tools are normally used. Right now, the absence of a predictable flat-fee option for solo users feels like a major gap.

A flat monthly AI option for individual users would make far more sense than forcing people into an unpredictable credit system for normal use. I understand charging more for heavy enterprise usage, large teams, or unusually high-volume workflows. But for independent professionals, job seekers, freelancers, and contractors, the current structure makes Make feel financially risky to rely on, even while already paying for a Professional seat.

What makes this especially frustrating is that other subscription products manage to absorb variation in individual usage without turning normal engagement into a stressful meter. Some people use a service lightly, others use it heavily, and the flat monthly price is part of what makes the product feel usable and trustworthy. For a creative assistant built into a design platform, that kind of predictability matters.

I’d strongly encourage Figma to consider changes such as:

a flat monthly AI option for individual Professional users
more generous included credits for active Make users
clearer per-action cost visibility before a request is sent
lower-cost or free handling for retries, corrections, and failed outputs
some form of session-based or workflow-based pricing rather than charging every conversational turn so aggressively
better protection against burning credits on normal back-and-forth required to make the tool usable

I’m sharing this because I like the product and want it to succeed. Make has been one of the most useful parts of Figma for me recently. But the current credit model makes me hesitant to rely on it, and that is a bad outcome for both users and the product.

@AaronUX well said, I hope you are heard by Figma team

Show more replies

Sign up

Login with SSO

Login to the community

Login with SSO