Agent Payments Protocol (AP2)

This repository contains code samples and demos of the Agent Payments Protocol.

Intro to AP2 Video

AP2 on The Agent Factory - September 2025

About the Samples

These samples use Agent Development Kit (ADK) and Gemini 3.1 Flash Lite Preview.

The Agent Payments Protocol doesn't require the use of either. While these were used in the samples, you're free to use any tools you prefer to build your agents.

Navigating the Repository

The top-level layout is:

• docs/ — specification, flows, FAQ, and MkDocs sources.
• code/ — all source code, organized by artifact:
  • code/sdk/ — the AP2 SDK (Python lives at code/sdk/python/ap2/).
  • code/samples/ — reference implementations and scenarios.
  • code/web-client/ — the demo web client (Vite + React).

The samples directory contains a collection of curated scenarios meant to demonstrate the key components of the Agent Payments Protocol.

The scenarios can be found in:

• code/samples/python/scenarios — Python.
• code/samples/go/scenarios — Go.
• code/samples/android/scenarios — Android.

Each scenario contains:

• a README.md file describing the scenario and instructions for running it.
• a run.sh script to simplify the process of running the scenario locally.

This demonstration features various agents and servers, with most source code located in code/samples/python/src. Scenarios that use an Android app as the shopping assistant have their source code in code/samples/android; Go roles live in code/samples/go.

For the SDK API reference see code/sdk/python/ap2/sdk/README.md.

Quickstart

Prerequisites

• Python 3.11 or higher
• uv package manager

Setup

You can authenticate using either a Google API Key or Vertex AI.

For either method, you can set the required credentials as environment variables in your shell or place them in a .env file at the root of your project.

Option 1: Google API Key (Recommended for development)

1. Obtain a Google API key from Google AI Studio.

2. Set the GOOGLE_API_KEY environment variable.

   • As an environment variable:

     export GOOGLE_API_KEY='your_key'

   • In a .env file:

     GOOGLE_API_KEY='your_key'

Option 2: Vertex AI (Recommended for production)

1. Configure your environment to use Vertex AI.

   • As environment variables:

     export GOOGLE_GENAI_USE_VERTEXAI=true
     export GOOGLE_CLOUD_PROJECT='your-project-id'
     export GOOGLE_CLOUD_LOCATION='global' # or your preferred region

   • In a .env file:

     GOOGLE_GENAI_USE_VERTEXAI=true
     GOOGLE_CLOUD_PROJECT='your-project-id'
     GOOGLE_CLOUD_LOCATION='global'

2. Authenticate your application.

   • Using the gcloud CLI:

     gcloud auth application-default login

   • Using a Service Account:

     export GOOGLE_APPLICATION_CREDENTIALS='/path/to/your/service-account-key.json'

How to Run a Scenario

To run a specific scenario, follow the instructions in its README.md. It will generally follow this pattern:

1. Navigate to the root of the repository.

   cd AP2

2. Run the run script to install dependencies & start the agents. The exact path depends on the scenario — for example, the human-present card payment flow:

   bash code/samples/python/scenarios/a2a/human-present/cards/run.sh

   Other scenarios live alongside it under code/samples/python/scenarios/, code/samples/go/scenarios/, and code/samples/android/scenarios/ (see each scenario's README.md for the exact invocation).

3. Navigate to the Shopping Agent URL and begin engaging.

Installing the AP2 Types Package

The protocol's core objects are defined under code/sdk/python/ap2/ — Pydantic models in models/ and sdk/generated/, canonical JSON schemas in schemas/. A PyPI package will be published at a later time. Until then, you can install the package directly using this command:

uv pip install git+https://github.com/google-agentic-commerce/AP2.git@main