Agents

Google ADK

Google's open-source Agent Development Kit for building multi-agent systems with Python, TypeScript, Go, and Java.

20K Stars
Apache-2.0 License
google Maintainer
2026-06-16 Verified
Overview

What is Google ADK?

Google's open-source Agent Development Kit for building multi-agent systems with Python, TypeScript, Go, and Java.

FAQ

Frequently asked questions

What should I check before using Google ADK?

Build one two-step agent flow with a tool call, a state transition, and an observable failure path.

Is Google ADK open source?

Google ADK is listed on OpenAgent.bot with Apache-2.0 based on the current resource metadata. Re-check the official repository, docs, and license before production use.

Decision brief

Should you use Google ADK?

JSON
Best for
  • Builders evaluating this category
Not for
  • Production adoption without checking source links, permissions, and maintenance.
Trust and freshness
  • Verified 2026-06-16
  • License: Apache-2.0
  • Repo: google/adk-python
  • Open-source signal
Deployment

Check source

Permission surface

Low explicit permission surface in metadata

Decision signals

Local first, API, CLI

Agent packet

Structured decision data for Google ADK

This packet is the compact machine-readable view agents should use before following source links or taking action.

Capabilities

workflow orchestration, tool calling, automation

Constraints

open source

Deployment

Check source

Permission surface

Low explicit permission surface in metadata

Recommended workflows

developer workflow, production agent

Overview

What Google ADK does

What it is

Google ADK is a agent in the agents category. Google's open-source Agent Development Kit for building multi-agent systems with Python, TypeScript, Go, and Java.

Why it matters

Google ADK matters when builders need a clearer way to choose tools by workflow fit, constraints, source quality, and operational risk rather than by category labels alone.

How to evaluate it

Evaluate Google ADK by starting from the official sources, checking its interface surface, and running one narrow workflow before expanding scope.

Facts

Known metadata and operating surface

These fields are separated from editorial interpretation so agents can reason over facts and missing checks.

Resource type agent
Category Agents
Maturity active
Difficulty Unknown
License Apache-2.0
Pricing open source
Verified 2026-06-16
Source confidence medium
Risk level moderate
Fit matrix

Where Google ADK fits in an agent stack

partial

Browser automation

Google ADK has at least one signal for browser automation, but should be checked against a real task before adoption.

  • Run one non-sensitive website task and inspect clicks, waits, retries, and changed URLs.
  • Confirm official docs, current maintenance, license, and runtime constraints before production use.
partial

Coding agent workflow

Google ADK has at least one signal for coding agent workflow, but should be checked against a real task before adoption.

  • Run a small repository change and inspect the diff, tests, and rollback path.
  • Confirm official docs, current maintenance, license, and runtime constraints before production use.
partial

Reusable skill workflow

Google ADK has at least one signal for reusable skill workflow, but should be checked against a real task before adoption.

  • Run one skill end to end and check whether it produces evidence or structured output.
  • Confirm official docs, current maintenance, license, and runtime constraints before production use.
weak

Connector or protocol layer

Google ADK is not primarily positioned for connector or protocol layer in the current metadata.

  • Connect one low-risk service, then inspect schemas, auth scope, errors, and logs.
  • Confirm official docs, current maintenance, license, and runtime constraints before production use.
weak

Evaluation and observability

Google ADK is not primarily positioned for evaluation and observability in the current metadata.

  • Add one repeatable test case and confirm results can run again in review or CI.
  • Confirm official docs, current maintenance, license, and runtime constraints before production use.
weak

Local or private AI stack

Google ADK is not primarily positioned for local or private ai stack in the current metadata.

  • Verify hardware requirements, data path, storage, and whether all calls stay in your environment.
  • Confirm official docs, current maintenance, license, and runtime constraints before production use.
Inputs and outputs

What an agent should inspect

Likely inputs

  • Official setup instructions and a small real workflow

Likely outputs

  • A decision on whether this resource fits the target workflow
Evidence

Sources, claims, and missing checks

Claims are marked separately from source links so future crawlers and reviewers can update them without rewriting the page.

verified

Google ADK is listed as open source.

License metadata: Apache-2.0
verified

Google ADK has a recorded GitHub repository: google/adk-python.

Resource facts and GitHub source link.
Missing checks
  • Dedicated docs link is missing.
  • Repository freshness has not been recorded.
Next action

How to start evaluating Google ADK

Inspect repository

Check license, recent activity, issues, examples, and security-sensitive code paths.

Open source
Compare

Alternatives and nearby resources

Use related resources to compare category fit, license, deployment model, and first-workflow behavior.

FAQ

Common questions about Google ADK

What is Google ADK used for?

Google ADK is used as a agent for agents workflows. The most relevant recorded capabilities are workflow orchestration, tool calling, automation.

Is Google ADK open source?

Google ADK is listed as open source with Apache-2.0 license metadata. Re-check the official repository or source link before production use.

Can agents use Google ADK directly?

Google ADK has recorded interfaces such as official source links. Agents should prefer the JSON or Markdown profile first, then follow official docs for real execution.

What should I check before production use?

Check source confidence (medium), risk level (moderate), license, maintenance freshness, permission surface, required credentials, and whether the first workflow succeeds in a sandbox.