Agent Development Kit

The framework for agents that actually ship

Open-source, Python-first, with built-in tools for testing, debugging, and deployment. From prototype to production in one framework.

Get started GitHub →
$ pip install google-adk

Your coding agent already knows ADK

ADK ships as a Skill and MCP server — coding agents like Claude Code and Gemini CLI can scaffold, extend, and test your agents automatically.

$ claude "Build me a weather agent with ADK"
⠋ Using ADK skill + MCP server...

from google.adk import Agent
from google.adk.tools import google_search

agent = Agent(
  name="weather",
  model="gemini-2.5-flash",
  tools=[google_search],
)

Agent created with tool bindings
🌐

Any Model, Your Infra

Use Gemini, GPT, Claude, or open models. Run anywhere — your laptop, Cloud Run, or your own infrastructure.

📈

Low Floor, High Ceiling

Start with a single agent in 5 lines. Evolve into multi-agent graphs with deterministic control flow when you need it.

ADK 2.0 — Graphs + Code

Graph-based agents combine AI reasoning with deterministic logic. Mix LLM calls with regular code for workflows you can trust.

Built-in Dev UI

Interactive web UI for testing, debugging, and tracing your agents. See every tool call, inspect state, and iterate fast.

ADK Dev UI
Run adk web to launch. Test conversations, inspect tool calls, trace agent behavior in real time.

Go beyond vibes

ADK's eval framework tests both the final response AND the agent's trajectory — the steps it took to get there.