Ten changes to how Figma explains what Make costs

Ten changes, in order

1. Put the add-on’s prices on the pricing page

   Seen in Fig. 3. Show it on the figure.no price on the cardrev. 20.10

   Seen, why, try

   Seen. The pricing page’s “AI credits” add-on, “Purchase additional AI credits for your whole team to share—enable pay-as-you-go for flexibility and/or subscribe to save”, shows no price. Its “Learn more” leads to the help article “Manage AI credits”, where the pay-as-you-go rate is shown “from the Add credits flow in your admin dashboard”. The AI credit updates FAQ says the same (Fig. 3). No public page I read prices either option.

   Why it matters. Figma’s help says seat credits “allow you to try Figma AI and see what it can do”, and “For ongoing, everyday use across your team, your admin can add more credits” (the inset). So everyday use of Make has the one cost a buyer cannot see before buying. The help article on buying credits is marked “Available on paid plans” and “Admins only” (the inset under change 08): whoever approves the spend learns the price once the team is on a paid plan, or from sales.

   Try. A short table under the plan cards: each subscription tier (credits and price a month), the pay-as-you-go rate per 1,000 credits, and the date of the prices. “Subscribe to save” then has figures behind it.

   , open it full size

2. Tell Make’s own visitors what Make runs on

   Seen in Fig. 1. Show it on the figure.free, with how many credits?

   Seen, why, try

   Seen. figma.com/make is 11,078 px long at 1440 px and never uses the words credit, seat or price; the way to a price is the Pricing link in the menu and the footer. Its button says “Get started for free” (Fig. 1).

   Why it matters. The free start is real: the free Starter plan includes 150 AI credits a day, up to 500 a month. But a visitor who reads only this page learns what Make does and nothing about what it uses up. By Figma’s own example a new app from scratch takes about 100 credits or more, so a free day covers one and not two (my arithmetic).

   Try. One line under the prompt box: “Make runs on AI credits: 150 a day on the free plan, 3,000 a month with a Professional Full seat. A new app takes about 100 or more.” Then a link to the credits on the pricing page.

3. Replace “100+” with a typical figure and a ceiling

   Seen in Fig. 4. Show it on the figure.a floor, with no ceiling

   Seen, why, try

   Seen. The help article gives three Make examples: changing the font, “Approximately 30+ credits”; making an attached design interactive, 75+; generating an app from scratch, 100+ (Fig. 4). Each is a floor. The same article says that your own usage history “will provide the most accurate estimate”.

   Why it matters. A floor tells a buyer the least a task can cost, not what it usually costs. Against a Professional Full seat’s 3,000 credits a month, “100+” means at most 30 apps from scratch, perhaps far fewer, and the page cannot say which. The usage history the article points to already exists: Make shows each prompt’s cost once it has run.

   Try. For each example, the median and the 90th percentile from real use, with the model and the month: “Generate an app from scratch: typically [median], nine in ten under [90th percentile]. Auto, October 2026.”

4. Name the default model, and date the examples to it

   Seen in Fig. 5. Show it on the figure.two names for the defaultrev. 20.10

   Seen, why, try

   Seen. The Make examples are “calculated using our default model as of February of 2026”, under a rate table dated “as of October 15, 2026” (Fig. 4). The help page on models lists six models “in addition to the default model”. Figma’s guide to saving credits in Make calls it “Auto (default)” (Fig. 5).

   Why it matters. The only figures for Make are about eight months older than the rates beside them, measured on a default that one page calls “the default model” and another “Auto”, and no page says which model that was in February. A buyer cannot tell whether they still hold.

   Try. Say in one sentence what Auto is (one model, or a choice among several for each prompt), use that one name everywhere, and measure the examples again whenever the rates are dated.

5. Put a relative cost beside each model

   Seen in Fig. 5. Show it on the figure.six models, and no figure for any of themrev. 20.10

   Seen, why, try

   Seen. Make’s model list names GPT‑5.6 Terra, GPT‑6.1 Sol, Claude Sonnet 4.5, Claude Opus 5.5, Gemini 3.8 Flash and Gemini 4 Argon: models from OpenAI, Anthropic and Google. On cost it gives no figures: credit consumption for these models “is still in development and may vary significantly”, and some “use substantially more credits per task” (Fig. 5). Figma’s guide to saving credits calls Gemini Flash “the most credit-efficient option for straightforward tasks” (the inset).

   Why it matters. The model is chosen in the chat box, as the user works, and “substantially more” could mean twice as much or ten times. Figma’s article on how credits work already prints a fixed cost per model for the image tools: 2, 6, 8 or 16 credits an image.

   Try. A multiplier against Auto for each model, from the same usage data as change 3 and dated like the rates, in the help page and beside each name in the picker.

   , open it full size

6. Show what a prompt is likely to cost before it runs

   Seen in Fig. 6. Show it on the figure.shown only after it runsrev. 20.10

   Seen, why, try

   Seen. The help article asks “Can I see the cost of a prompt in Figma Make or with the agent before I run it?” Its answer: “we can’t predict exactly how many credits it will use”, and the cost is shown afterwards, by hovering over the credits icon under the finished prompt (Fig. 6). I found nothing on the pages I read about credits running out while a prompt runs.

   Why it matters. On a free day of 150 credits, one prompt of 100 or more can use most of the day, and the user learns that once it has been spent. Someone who cannot see the price of a prompt has a reason to send fewer of them.

   Try. Before a prompt is sent, a range drawn from similar prompts (“usually between [low] and [high] credits”) and an optional ceiling (“stop and ask me above [n]”). In the docs, one sentence on what happens when credits run out partway through a prompt.

7. Name Make in words on the seats that include it

   Seen in Fig. 2. Show it on the figure.Make: one icon in tenrev. 20.10

   Seen, why, try

   Seen. On the plan cards each seat has a row of 16 px product icons and a credit line. Make’s icon is the second of ten on Full seats and is absent from Dev and Collab seats, which still show “+ 500 AI credits/mo” (Fig. 2). The words are further down, in the FAQ answer to “What is a seat?” (the inset).

   Why it matters. A buyer choosing seats for Make has to match a 16 px icon to the row of product names above the cards, and can easily read a Dev or Collab seat’s 500 credits as credits for Make.

   Try. Put it in words on the card: “Full seat: every product, Make included”; on Dev and Collab seats, “500 AI credits a month, for the products in this seat (not Make)”.

   , open it full size

8. Make the banner match the rule

   Seen in Fig. 2. Show it on the figure.all plans, or every paid plan?rev. 20.10

   Seen, why, try

   Seen. Above the plan cards, under a “NEW” badge: “AI add-ons are now available for all plans.” (Fig. 2). The help article on buying credits is marked “Available on paid plans” and “Admins only” (the inset), and Starter, the free plan, is one of the four plans on the page.

   Why it matters. A Starter user who reads the banner, uses up the day’s 150 credits and looks for the add-on will find it is not offered to them.

   Try. “AI credit add-ons are now available on every paid plan.” And on Starter’s card, the step up in one line: “Need more? A Professional Full seat includes 3,000 a month.”

   , open it full size

9. Unfold the Make examples, and link straight to them

   Seen in Fig. 4. Show it on the figure.folded: a reader has to know to open it

   Seen, why, try

   Seen. The Make examples sit in a folded panel, 5,088 px down an article the page labels “18 min” (Fig. 4). The pricing page’s FAQ on credits links to the top of that article.

   Why it matters. These three lines are the nearest thing on any public page to a price for a Make prompt, and a buyer who does not scroll that far, or does not think to open the panel, never sees them.

   Try. Open the panel by default, give it an address of its own (“#figma-make”), and link to it from the pricing page’s credits FAQ and from Make’s own page.

10. On a phone, show Starter’s credits on the card

    Seen in Fig. 7. Show it on the figure.Starter’s credits, folded awayrev. 20.10

    Seen, why, try

    Seen. At 390 px each paid seat shows its credits on the card; Professional’s “+ 3,000 AI credits/mo” is 1,159 px down the page. Starter’s “150 AI credits/day, up to 500 AI credits/mo”, shown openly at 1440 px, is folded inside “Why choose Starter?”, below a first screen of 844 px (Fig. 7). The pricing FAQ gives Starter’s 500 a month and leaves out the 150 a day.

    Why it matters. On a phone the free plan is the only one whose AI allowance is out of sight, and it is the one with a daily limit.

    Try. One line under “Free”: “150 AI credits a day, up to 500 a month”, and the same daily limit in the FAQ answer.