OpenUI

Overview

OpenUI is the open standard for generative UI. It enables AI apps and agents to respond with interactive UI, charts, tables, forms, and dashboards. OpenUI Lang is a compact, streaming-first language that LLMs write to build interfaces from your own components, and it is not JSON.

Key Features

• OpenUI Lang is a compact, streaming-first language that LLMs write to build interfaces from your own components; each line defines one component, such as root = Stack([header, chart]).
• Interfaces render while the response is still streaming.
• Works with any model, any frontend framework, and any design system or component library you already use.
• Can generate interactive UI, charts, cards, dashboards, forms, and tables.
• The benchmark page states OpenUI Lang generates up to 3x faster and uses 67% fewer tokens than JSON-based UI formats.
• OpenUI is fully open source and works with your own components and design system.

Use Cases

• Build AI apps and agents that respond with interactive UI, charts, tables, forms, and dashboards.
• Render dashboards, forms, tables, charts, and complete application screens from a response, not only chat messages.
• Add OpenUI to existing agent code using integrations such as Vercel AI SDK, LangChain, LangGraph, Convex, Cloudflare, CrewAI, n8n, and Haystack without a rewrite.
• Work within existing frontend frameworks such as React, Next.js, Vue, Svelte, Angular, Nuxt, Remix, and Astro.

Getting Started

• Run pnpx @openuidev/cli@latest create to scaffold a working app.
• Run npx @openuidev/cli@latest setup-agent in your project to add OpenUI instructions to AGENTS.md for coding agents such as Codex, Cursor, and Claude Code.
• Projects created with npx @openuidev/cli@latest create already include the agent instructions.

Before You Adopt

• License: MIT. Review its terms before using, modifying, or distributing the project.
• Production deployments may need OpenUI Gateway to catch and repair invalid generated output before it reaches users.
• Gateway can switch model providers during an outage while keeping the same model, which is an operational reliability consideration.