Generative UI
Generative UI is an interface produced by a model from a prompt — a whole screen, flow, or prototype generated rather than assembled by hand.
Generative UI means a model produces the interface itself — a screen, a flow, or a working prototype — from a prompt and whatever design system it was given. The useful version is grounded in your components; the useless version invents a new visual language every session.
Generative UI is an interface produced by a model rather than assembled by hand: you describe a screen, a flow, or a whole prototype, and the tool builds it, usually as real code you can click through. The category grew out of prompt-to-app tools and now covers everything from a single component to a multi-screen prototype with working state.
Three things separate a useful generative UI tool from a demo. First, grounding: does it read your design system, or invent one? Claude Design ingests a system from a repository, design files, or uploads and reuses those colors, typography, and components on every later project. Second, editability: can you fix the output where it is wrong, by editing the canvas, the code, or both? Third, exit: v0 generates React and Tailwind you can sync to GitHub and deploy, which matters when the prototype is meant to become the product. Claude Design vs Figma Make compares two of the main options on exactly those axes.
Generative UI runs in the opposite direction from design-to-code, which starts from a design that already exists and translates it. In practice most teams use both: generate to explore, then translate the version everyone agreed on into the codebase. The Claude Design guide is the workflow version of that loop, and the designer tools roundup covers who does which.
Frequently asked questions
- Is generative UI the same as vibe coding?
- They overlap but are not the same. Vibe coding describes accepting model-written code without reading it closely. Generative UI describes the output: an interface produced by a model. You can generate a UI and still review every line, and you can vibe code something that is not an interface at all.
- Does generative UI produce throwaway work?
- It depends on what you feed it. A generator with no design system produces a plausible-looking screen in a house style that is not yours, which is exploration at best. A generator that has ingested your tokens and components produces something a developer can actually start from.
- Where does the designer's job go?
- Upstream and downstream. Upstream is defining the system, the constraints, and the brief the generator works from. Downstream is judging the output, editing it, and deciding what ships. The middle — drawing the fifth variant of a settings screen by hand — is the part that compresses.
Related
- Design-to-CodeDesign-to-code is turning a design — a Figma file, a screenshot, or a generated mockup — into working front-end code, increasingly by an AI agent.
- Claude Design vs Figma Make: Which Prompt-to-Prototype Tool?Claude Design vs Figma Make compared on what each makes, plan and seat requirements, design-system input, refinement, export, and handoff to code.
- The Best AI Tools for Designers in 2026The AI tools designers actually use in 2026: prototyping, websites, image generation, decks, design-to-code, and design systems, with a verdict per category.
- Claude DesignAn Anthropic Labs tool for making prototypes, decks, and one-pagers with Claude, using your design system, with PPTX, PDF, HTML, and Canva export.
- Claude Design: The Complete Guide (2026)What Claude Design makes, how it ingests your design system, the three editing modes, every export and handoff, and where it stops. Written for designers.
- V0Vercel's generative UI builder that turns prompts into production-ready React, Next.js, and shadcn/ui apps.