Figma acquired the team behind Bud, a Y Combinator-backed vibe-coding platform previously known as Orchids. The products shut down on July 18, according to TechCrunch. The deal is structured as an acquihire: only the engineers move, the apps do not.
Financial terms, headcount, and where the engineers will sit inside Figma have not been disclosed. The only public statement of intent came from Bud CEO Kevin Lu on X. Figma, he wrote, is "where ideas start, iterate, and come to life." His post is the entire deal rationale on the record. Figma's newsroom entry adds no specifics.
Bud originated as a vibe-coding app builder for mobile, web, Slack, and browser surfaces, then rebranded into an agent platform that browses the web and writes code on a user's behalf. The defining feature was speed: minutes from prompt to working app, with the AI writing the code rather than a designer hand-stitching it. That cadence assumes a tolerance for breakage and a feedback loop measured in hours, not sprints. Figma's own design tooling runs at quarterly cadence by comparison.
Figma is using the deal to buy a capability it could not build at the same tempo: consumer-grade prompt-to-prototype speed. Recent launches on the company's blog push the product from pure design toward building and prototyping. TechCrunch's reporting paraphrases the stated direction as making Figma "more than a design platform" by pulling the coding and prototyping layer closer to the design canvas. The Bud team fills that gap.
Cursor, Replit, and Windsurf have made AI-written code inside the editor the default developer expectation. Figma's core users sit one step earlier, in the no-code-yet phase where there is only a Figma file. That is a moat if Figma can connect it: every team that already mocks up interfaces in Figma is one AI step away from a working app, and the seat count lives upstream of any Cursor logo. By absorbing the Bud team, Figma is trying to own the prompt-to-prototype moment rather than hand it off.
Figma keeps the design canvas and brings the messy, fast vibe-coding layer in-house through the acquihire. Cursor and Replit own the editor. Windsurf owns the IDE-with-agent lane. Figma wants the no-code-yet layer above all of them.
Figma's deal fits the standard acquihire template: the buyer keeps the platform, the acquirer keeps the engineers. Terms are undisclosed. Standard carve-outs also specify which prior IP moves with the team and which stays behind. That inventory question is open. Pace, not code, is the asset Figma is importing.
Earlier in 2026, the BBC reported, via a security researcher, that apps generated on Orchids were vulnerable to cyberattacks. The report was secondhand at the time and the issue remains unresolved. Figma-bound work is not the same surface as Bud or Orchids, but review discipline travels with the engineers rather than the products.
The July 18 shutdown is the next concrete trigger. The prototypes that come out of the Figma canvas in the back half of the year will be the first signal of whether the team kept its pace inside a public-company cadence.