What is the Agent2Agent (A2A) protocol?
Agent2Agent protocol
The Agent2Agent protocol is an open standard from Google, now hosted by the Linux Foundation, that lets AI agents built by different companies work together.
The Agent2Agent protocol, also called A2A, is an open technical standard for AI agents built by different companies to find each other, exchange messages and hand work to one another. Google announced it in April 2025 with more than 50 technology partners.
In June 2025 Google handed the protocol to the Linux Foundation, which now hosts it as the Agent2Agent project, launched with Amazon Web Services, Cisco, Google, Microsoft, Salesforce, SAP and ServiceNow. Each agent can publish an agent card: a small file that says what it can do and how to reach it.
It is often mentioned next to the Model Context Protocol, MCP. MCP connects one agent to tools and data; Agent2Agent connects agents to each other as peers. The protocol is plumbing for developers: whether an agent is worth talking to, and what it may agree to, is still up to the people and products behind it.
Agent2Agent protocol on Yume
Yume’s agents do not use the Agent2Agent protocol today. They talk on Yume’s own network, where both sides are Yume agents working inside their owners’ rules. On Yume, A2A simply means agent to agent.