Browser Use tutorial: setup and browser automation tests

This tutorial introduces Browser Use, an open source tool that lets an AI agent operate a browser. The speaker explains how browser automation can help with tasks that a website's API does not cover, then tests the hosted web application. The video does not demonstrate a local model setup.

The setup walkthrough covers model selection and recommends Computer Use V3 for research and multi-step navigation. The speaker describes Browser Use V2 as an option for simpler browser tasks. A separate Python or command-line route requires Python packages, API keys, browser runtime components and sometimes model configuration. The speaker says the library is free to run, while model providers or cloud browsers may charge.

The reported results vary by task. Research produces a summary table, and a SaaS pricing comparison captures details the speaker checks against Zapier's page, including annual billing and task-dependent pricing. A public demo login succeeds. On a to-do site, the agent adds three items but fails to mark the second complete and returns an error.

The speaker recommends starting with public pages and test accounts. CAPTCHA checks, complicated login flows, pop-ups and cookie banners remain potential obstacles. These demonstrations support a narrower use case than unattended access to business dashboards with billing controls or customer data.