# Vercel Launches v0 API for Headless App Building > Source: > Published: 2026-08-13 15:37:00+00:00 Vercel has made the [v0 API](https://vercel.com/blog/introducing-the-new-v0-api) generally available, providing programmatic access to its AI-powered application-building agent. Developers can send prompts through the API, have v0 generate and modify application files, run the resulting app in a [Vercel Sandbox](https://vercel.com/docs/sandbox), and receive a preview URL that can be embedded into their own interfaces. The API organizes each application around a chat that maintains the app's current state. Developers can send follow-up messages using the same chat ID, allowing [v0](https://v0.app/) to continue editing the existing application. The API supports synchronous, asynchronous, and streaming requests, with streamed responses exposing individual agent actions such as file reads, edits, searches, Bash commands, and tool calls. Applications can also be created from existing GitHub repositories, ZIP archives, or sets of files. When the app is ready, developers can attach a Vercel project and deploy it through the API. Preview environments use short-lived tokens and can be accessed through a server-side proxy, keeping the [v0 API](https://v0.app/docs/api/v2) key out of the browser. The API also supports external tools and design systems. Developers can connect [MCP servers](https://en.wikipedia.org/wiki/Model_Context_Protocol) or load skills from team or user memory, [skills.sh](http://skills.sh), or a connected repository. Up to three skills can be passed to a request. Vercel also provides integrations for using v0 as a tool inside other AI agents through MCP, the [AI SDK](https://ai-sdk.dev/), or Vercel's eve framework. This positions v0 differently from application-building tools such as [Lovable](https://lovable.dev/) and [Bolt](https://bolt.new/), which primarily provide interactive environments for generating applications. The [v0 API](https://vercel.com/blog/introducing-the-new-v0-api) instead exposes the application-building agent as infrastructure that developers can invoke from their own products, scripts, CI pipelines, or other agents. Vercel's approach also integrates the generated application directly with its Sandbox and deployment APIs. Community reaction has focused on the potential for the API to support fully programmatic application development. AI influencer [ZenithAI](https://x.com/ZenithAiLab/status/2085424422186635674) described the release as: A powerful step toward fully programmatic app development workflows. Vercel developer [@EstebanSuarez](https://x.com/EstebanSuarez/status/2085370198819815567) highlighted the intended agent workflow, explaining that an agent can send a prompt and design system to the v0 API, receive a live preview in a Vercel Sandbox, iterate on the application, and deploy it when ready, posting: Your agent needs an app. It sends a prompt to the v0 API along with your design system. Back comes a preview URL: the app live in a Vercel Sandbox. Your agent iterates on it. When it's ready, it deploys. That's the new v0 API. The new API is available through the v0 SDK, with the v2 API using https://api.v0.dev/v2. Developers can create an API key, install the SDK, and build interfaces that send prompts, stream agent activity, render previews, and deploy generated applications.