Explore our GitHub repository https://github.com/nottelabs/notte and star us 🌟
In this quickstart guide, we’ll create a simple web agent that can browse and interact with websites autonomously. You’ll learn how to set up your development environment, create a browser session, and run your first agent task.

Guide

1

Get an API key from the console

Create an account on our console to get an API key. We offer a free plan for testing and a Pro plan for production, with Enterprise options available.
2

Install prerequisites

You’ll need Python 3.11 or above and install our Notte Python SDK.
uv venv --python 3.11
source .venv/bin/activate
uv pip install notte-sdk
3

Run your first agent

The following script creates a session, attaches an agent to the session, and runs a basic task. Both session and agent have many parameters you can tweak.
from notte_sdk import NotteClient

client = NotteClient(api_key="your-key-here")

# create a browser session resource
with client.Session(headless=False) as session:

    # attach an agent to session
    agent = client.Agent(
      session=session,
      reasoning_model="gemini/gemini-2.0-flash",
      max_steps=5,
    )

    # run the agent on a task
    response = agent.run(
      task="doom scroll cat memes on google images"
    )

Next steps

Congratulations! This simple example demonstrates the core workflow - creating a session, attaching an agent, and giving it a task to execute. From here, you can explore more advanced features like custom reasoning models, session replay, and integrations with other AI tools. Check out our other guides to learn more about Notte.