v0

v0 is Vercel's answer to a specific question: what if the starting point for any frontend component was a prompt instead of a blank file? It's not trying to build your whole app. It's trying to get you to working React faster than you would get there yourself.

The tool generates React components using Tailwind CSS and shadcn/ui. If you know those three things, you already understand what v0 is good at and where it stops. The output follows current React conventions, imports from the right places, and doesn't do anything structurally weird. It's scaffolding that you'd actually use rather than delete and rewrite.

Vercel shipped a significant rebrand and upgrade in January 2026: v0.app replaces the previous interface, adds a full-stack sandbox with Git integration and database connectors to Snowflake and AWS. It's moved from component generator toward a more complete development environment, though the component generation is still its clearest strength.

The Figma import is worth noting. You can drop in a frame from Figma and v0 will produce the React component equivalent. It's not pixel-perfect and it doesn't understand complex component states automatically, but for straightforward UI (a card, a form, a navigation bar) it saves the transcription work.

The model behind v0 (v0-1.5-lg with a 512K context window) handles reasonably complex component requests: multi-state forms, data table variants, responsive layouts. Where it predictably falls short is anything requiring real application logic; it generates UI structure, not business logic. Don't expect it to understand your authentication flow or wire up your API endpoints without significant hand-holding.

Who v0 is for: frontend developers on the Vercel/Next.js stack who want to move faster on UI scaffolding. Design engineers who think in React and want to prototype component ideas quickly. Anyone already using shadcn/ui who wants AI assistance that actually understands their stack.

Who it's not for: designers who don't write code and are hoping for a no-code experience (there isn't one here), developers working outside the React ecosystem (v0's value proposition dissolves if you're writing Vue or Svelte), or teams expecting a complete app builder. Compare it with Bolt.new or Lovable if you want something that builds apps rather than components.

The billing is token-based, which is both a feature and a frustration. Five dollars a month in free credits is enough to get a feel for it. At $20 a month for Premium, the cost is defensible if you're generating components regularly. The unpredictability of token consumption on complex requests is a genuine issue; it's harder to budget for than a flat monthly seat.

The absence of collaboration features is conspicuous for a Vercel product. v0 is single-player. For a tool in the Vercel ecosystem, which is built around teams, you'd expect Vercel to fix this eventually.