AG-UI, or Agent-User Interaction Protocol, is an open event-based protocol for connecting AI agents to user-facing applications. It standardizes how agent state, messages, tool events, UI intents, and user interactions move between an agent backend and a frontend.
AG-UI addresses a different boundary than the Model Context Protocol (MCP) or the Agent2Agent Protocol (A2A). MCP connects agents to tools and data. A2A connects agents to other agents. AG-UI connects agents to the interactive application surface where a user can see progress, steer work, approve actions, or interact with generated UI.
Typical AG-UI use cases include:
- streaming model output and tool events into an application;
- rendering agent progress, plans, and intermediate work;
- exposing human-in-the-loop approvals;
- synchronizing shared state between frontend and agent runtime;
- supporting generative UI; and
- allowing users to interrupt, edit, retry, or redirect a long-running task.
The protocol is useful because agent applications are not simple request-response systems. Agents may run for minutes, call tools repeatedly, branch into sub-agents, and need user input before continuing. A stable interaction protocol makes those workflows easier to debug and reuse across frameworks.
Security and product design still matter. A client should make clear which agent is acting, what action is pending, which data is being shown, and whether a user approval changes external state.
The AG-UI documentation describes AG-UI as the event-based protocol for connecting agentic frontends with agentic backends.
The LLM Knowledge Base is a collection of bite-sized explanations for commonly used terms and abbreviations related to Large Language Models and Generative AI.
It's an educational resource that helps you stay up-to-date with the latest developments in AI research and its applications.
Previous