What is AG-UI, the Agent-User Interaction protocol?
AG-UI (Agent-User Interaction protocol) is an open, lightweight, event-based protocol that standardises how AI agents connect to user-facing applications, so a front end can follow and respond to an agent as it works.
Checked
An agent that runs for a minute and calls six tools needs a different front end from a chatbot that returns one reply. The user should see progress, tool activity and intermediate results, and be able to step in. Without a shared protocol, every agent framework and every front end wires this up differently.
AG-UI defines that connection as a stream of events, so an agent built on one framework can drive a front end built for another.
Where AG-UI fits
The AG-UI documentation positions the protocol as the third layer beside two others: MCP connects agents to tools, A2A connects agents to other agents, and AG-UI connects agents to users. It grew out of CopilotKit's work with LangChain and CrewAI.
AG-UI is distinct from A2UI, a separate specification for generative UI widgets.
Status as of October 2026
AG-UI 1.0.0 shipped on npm and PyPI on 2026-09-17. Breaking changes in 1.0 include replacing THINKING_* events with REASONING_* events and moving the finish reason into metadata.
The AG-UI docs list support in LangChain, CrewAI, Microsoft Agent Framework, Google ADK, AWS Strands, Bedrock AgentCore, Mastra, Pydantic AI, Agno, LlamaIndex, AG2, the Claude Agent SDK and Cloudflare Agents, with the OpenAI Agents SDK marked in progress. AG-UI is not a project of the Linux Foundation's Agentic AI Foundation.
How Motiontic's engineers apply it
When an AI product needs an agent inside its own interface, Motiontic's engineers decide early how the user will see and steer the agent: what streams, what pauses for approval and what can be undone. AG-UI is one way to make that contract explicit, and it keeps the front end independent of the agent framework underneath.
Where each agent protocol sits
| Protocol | Connects | Example |
|---|---|---|
| MCP | Agent to tools and data | An agent reads a CRM record |
| A2A | Agent to agent | An agent delegates a quote request to a supplier's agent |
| AG-UI | Agent to user interface | An app shows an agent's progress and asks the user to approve a step |
Common questions
What is AG-UI?
AG-UI, the Agent-User Interaction protocol, is an open, event-based protocol for connecting AI agents to user-facing applications, so front ends can show and respond to what an agent is doing as it works.
How is AG-UI different from MCP and A2A?
MCP connects an agent to tools and data, A2A connects agents to each other, and AG-UI connects an agent to the application a person is using. The three address different boundaries and can be used together.
Which frameworks support AG-UI?
The AG-UI docs list LangChain, CrewAI, Microsoft Agent Framework, Google ADK, AWS Strands, Mastra, Pydantic AI, LlamaIndex, the Claude Agent SDK and others, with OpenAI Agents SDK support in progress as of October 2026.
Related
A2A protocol
The open protocol for agents talking to other agents, now under the Linux Foundation.
MCP vs A2A
The protocol for tools and data, compared with the protocol for agents talking to agents.
Human in the loop
Approval gates where a person decides, placed by risk rather than on every step.
AI product development
New products and major AI features, built from scratch alongside the client's team.
Running agents that have to hold up?
Describe the agent, the systems it touches and what it must never get wrong. The first call is with a lab engineer, who will say plainly whether it is ready for production, and whether Motiontic is the right team.