Figma Says AI Makes Design More Valuable as Code Gets Easier

Dylan Field framed Figma’s quarter as evidence that AI is not simply a threat to creative software vendors. His argument was that as large language models make code easier to generate, the scarce layer above code becomes design.

Figma reported revenue growth of 46% year over year, according to Field, with net dollar retention of 139% among customers above $10,000 in annual recurring revenue. He also pointed to 16% non-GAAP operating margin and 27% free cash flow for the quarter, alongside a raised full-year outlook.

Those numbers were offered in response to Caroline Hyde’s question about whether large language models and frontier AI companies, including Anthropic, are disrupting software providers. Field’s answer was not that AI is irrelevant to Figma’s business. It was that AI changes where value concentrates.

“As AI commoditizes code and makes it so that code is easier than ever to write,” Field said, “the layer above code, as that gets commoditized, is design.”

That view runs through the rest of his explanation. Figma’s position, as Field described it, depends on design becoming more important inside companies, not less. He argued that design is spreading beyond specialist design teams and becoming a broader way companies compete, shape user experience, and “break through” an increasingly crowded information environment.

Ed Ludlow focused on one of the more concrete business questions: Figma’s move in March to begin charging customers for AI usage beyond a certain limit. Ludlow described mixed customer responses: many users who hit the cap were willing to pay for more credits, while a smaller group appeared to pull back from using a Figma product if additional AI use required payment.

Field’s answer separated Figma’s traditional design product from its AI-intensive features. Free users can still use free services, and paid-seat users can use the traditional design tool. But users who want Figma Make, or AI features inside Figma Design, now operate within a credit model.

Field said Figma added a number of free credits to paid seats so customers could try the AI features. Customers who want more can buy additional credits. He acknowledged that the company had kept the features free for a long period, but said that was not sustainable indefinitely because inference “does cost real money.”

The product footage shown during this discussion illustrated the breadth of the AI surface Figma is trying to monetize. The interfaces included AI-generated geometric forms, website mockups such as gallery and nature-themed landing pages, node-based editing workflows, and product screens labeled “Figma Sites,” “Figma Make,” and “Figma Draw.” The on-screen examples showed AI being used not merely to produce isolated assets, but to generate and manipulate design systems, page concepts, visual styles, and structured outputs.

Field also highlighted Figma Weave, which he described as a node-based editing tool for connecting model outputs such as images, videos, 3D models, and other media. The goal, in his phrasing, is to let users “mold those model outputs like clay” through a workflow rather than accept a one-shot generation.

His example was NBBJ, an architecture firm and Figma customer mentioned on the earnings call. Field said the firm previously conducted extensive site shoots to understand lighting at different times and then superimposed 3D building models. With Figma Weave, he said, that process can be handled inside a workflow where the firm controls parameters more easily, saves time, and produces better client results. That example was also his bridge back to the cost question: this is another use case where AI inference becomes part of the product economics.

The central financial tension is that AI can expand Figma’s market while also pressuring margins through token and inference costs. Ludlow put the question directly, citing an audience question from a regular viewer, Ben: how will inference and token costs affect margins going forward?

Field said Figma had discussed the issue on its earnings call. His answer was that if the company sees an opportunity for large growth, it will pursue it aggressively, even if that creates short-term margin pressure.

He distinguished between the short term and the long term: some growth investments may weigh on margins now, but he argued that if they open a “massive TAM,” that is the move investors should want. He identified the opportunity as spanning design, “sculpting,” advertising, marketing, and the broader problem of breaking through noise.

That answer did not deny the cost issue. It treated inference cost as a strategic input rather than an accounting footnote. Figma’s AI products require paid usage because model output is costly, but Field argued that the same AI usage could enlarge the company’s addressable market if Figma becomes the place where more kinds of workers shape, refine, and deploy visual and interactive work.

The second product screenshot reinforced that direction. It showed Figma interfaces for brush effects, event flyers, text blur on the word “VOLUME,” vector adjustment of a butterfly shape, texture controls, and data-table population. The visible labels included “Brushes,” “Layer blur,” “Texture,” “NATURE FEST,” and scientific-name table fields. The point of the demonstration was practical breadth: image treatment, layout, typography, vector manipulation, structured content, and publishing-oriented design all appearing within the same product environment.

Hyde put the market-pressure question to Field in more human terms. She referred to anecdotal evidence that employees who rely on Figma love it intensely enough that removing it from the workplace would provoke revolt. But she asked how Field fights an investor narrative of “sell first, ask questions later,” which she said had pressured the stock since the IPO.

Field’s answer was that Figma can only control the inputs: deliver for customers and make long-term decisions. He argued that the company’s strategy is aligned with where the market is going because design is becoming more central to how companies win.

The clarification he wanted to make was definitional. Design, in his telling, is not simply making something beautiful, especially because aesthetics vary. It is “how it works.” It is user experience, form, function, and the thinking process that produces those outcomes.

That matters to his AI argument. If design is just surface polish, then AI-generated interfaces and assets might look like commoditization. If design is a broader process of deciding how products work, how workflows are shaped, and how companies communicate in a noisy market, then Figma’s bet is that AI increases the number of people participating in design rather than replacing the need for design tools.

Field said that is what Figma is seeing with customers: not just designers using the product, but many others entering the design process. The company’s task, as he described it, is both to bring more people in and “level them up on design.”