Build AI Agents.
Open & Scalable.
Start quickly, evolve your designs, and scale up to enterprise-level deployments. Model-agnostic, deployment-agnostic, and fully open source.
from google.adk import Agent, GraphAgent # Define specialized agents researcher = Agent(name="researcher", model="gemini-2.0") writer = Agent(name="writer", model="gemini-2.0") # Compose into deterministic workflows workflow = GraphAgent(name="pipeline") workflow.add_node(researcher) workflow.add_node(writer) workflow.add_edge("researcher", "writer") workflow.run({"topic": "Agentic UX"})
pip install google-adk
com.google.adk:google-adk:0.6.0
npm install @google/adk
go get google.golang.org/adk
Build agents with agents.
ADK is designed to be written by both humans and AI. Hook up your favorite coding assistant to our MCP Server and let it generate robust, tool-bound agents in seconds.
Define your skills, bind your tools, and let your IDE do the heavy lifting.
Visual Debugging & Tracing
Stop printing to stdout. ADK includes a powerful interactive web UI for testing, debugging, and tracing agent behavior locally.
Inspect tool calls, modify context windows on the fly, and visualize multi-agent graph executions with zero configuration.
Go beyond vibes. Evaluate everything.
Testing agents is notoriously hard. ADK's built-in evaluation framework lets you systematically test not just the final text response, but the entire execution trajectory.
Assert that specific tools were called, check the exact sequence of graph nodes, and ground outputs against real data.