Google Expands ADK with Go Language Support for AI Agent Developers
Agent Development Kit for Go

Image Courtesy: Google

Google announced the official release of ADK for Go, bringing the Agent Development Kit (ADK) to one of the world’s most popular programming languages. With this launch, Google continues its mission to meet developers where they are — enabling them to build intelligent, high-performance AI agents using the tools and languages they know best.

ADK is an open-source, code-first toolkit designed for developers who want full control over their AI agents. It moves the complexity of large language model (LLM) orchestration, behavior definition, and tool integration directly into code — allowing for robust debugging, reliable versioning, and deployment freedom across local and cloud environments.

The Agent Development Kit gives developers the flexibility to design, debug, and deploy advanced AI agents tightly integrated with Google Cloud services. By defining logic, orchestration, and tool use directly in code, developers gain fine-grained control, improved testability, and a seamless development workflow.

Introducing ADK for Go

With the release of ADK for Go, developers can now harness Go’s renowned concurrency, performance, and strong typing to build scalable, reliable agentic applications. The Go version aligns with ADK’s core philosophy — empowering builders to define, test, and deploy AI agents with the same precision and rigor as any other modern service.

As a powerful addition, ADK Go includes out-of-the-box support for 30+ databases via the MCP Toolbox for Databases, simplifying data integration for developers.

Key Features of ADK Go

ADK Go provides a consistent development experience across all supported languages, including Python and Java, offering the same powerful toolkit and ecosystem benefits:

  • Rich Tool Ecosystem: Integrate pre-built tools, custom functions, and OpenAPI specs with ease, while enjoying seamless connections across Google’s platform.

  • Code-First Development: Define agent logic, orchestration, and tool use directly in Go for complete control, flexibility, and traceability.

  • Modular Multi-Agent Systems: Build scalable systems by composing specialized agents into flexible, maintainable hierarchies.

  • Built-In Development UI: Test, evaluate, debug, and showcase your agents with a visual interface that speeds up the development cycle.

A2A Protocol Support: Enabling Multi-Agent Collaboration

In addition to the Go SDK, Google is introducing full support for the Agent2Agent (A2A) protocol in ADK Go. This innovation enables developers to create multi-agent systems where agents can communicate, collaborate, and coordinate complex tasks securely — without exposing internal states or proprietary logic.

Google has also contributed the A2A Go SDK to the open-source A2A Project, expanding the interoperability of multi-agent systems. Developers can explore the A2A protocol and its ecosystem of SDKs at the A2A Protocol website

Get Started with ADK for Go

Developers can start building with ADK for Go today:

With ADK for Go, Google brings performance, precision, and flexibility together — empowering developers to build the next generation of AI-powered applications with confidence.