Agentic AI is starting to connect to APIs. Here are the emerging AI agent-to-API standards you need to know about.

Nordic APIs's resource offers insights, tutorials, and resources for API developers and architects. Readers can learn about API design principles, security practices, and API management strategies. With articles, webinars, and industry reports, Nordic APIs provides  guidance and expertise for building and managing APIs that power modern applications and digital ecosystems.

Nordic APIs

Seven emerging AI agent-to-API standards are compared to help developers choose the right protocol for connecting AI agents to APIs. The standards include Anthropic's MCP for extending LLMs with data sources, Google's A2A for agent-to-agent communication, Cisco's ACP/AGP for creating an "Internet of Agents," Wildcard's agents.json built on OpenAPI, LangChain's Agent-Protocol for production deployment, and Agile Lab's detailed YAML-based specification. Each standard serves different use cases, from local tool integration to multi-agent ecosystems, with many being complementary rather than competing solutions.

Comparing 7 AI Agent-to-API Standards

Cisco’s AGNTCY Agent Connect Protocol (ACP) and Agent Gateway Protocol (AGP) proposals

Final Thoughts on AI Agent-to-API Standards