A browser agent drives a real web browser — navigating, clicking, typing and reading pages — to complete tasks on sites that have no API.
2 min readupdated 2026-08-01
/ quick answer
Browser agents combine a vision or DOM-aware model with a browser automation layer (Playwright, Chromium DevTools Protocol). The model receives a page representation, decides the next action, and the runtime executes it. They unlock the long tail of systems without APIs, at the cost of speed, flakiness on layout changes, and detection by anti-bot systems.
A browser agent drives a real web browser — navigating, clicking, typing and reading pages — to complete tasks on sites that have no API. Browser agents combine a vision or DOM-aware model with a browser automation layer (Playwright, Chromium DevTools Protocol). The model receives a page representation, decides the next action, and the runtime executes it. They unlock the long tail of systems without APIs, at the cost of speed, flakiness on layout changes, and detection by anti-bot systems. In practice: An agent logs into a supplier portal weekly, downloads the invoice PDFs and files them in Drive — a portal with no API and no export. This dictionary node is part of the Onexial knowledge graph and links to related concepts, workflows and tools below.
Definition
Browser agents combine a vision or DOM-aware model with a browser automation layer (Playwright, Chromium DevTools Protocol). The model receives a page representation, decides the next action, and the runtime executes it. They unlock the long tail of systems without APIs, at the cost of speed, flakiness on layout changes, and detection by anti-bot systems.
Example
An agent logs into a supplier portal weekly, downloads the invoice PDFs and files them in Drive — a portal with no API and no export.
Browser agents combine a vision or DOM-aware model with a browser automation layer (Playwright, Chromium DevTools Protocol). The model receives a page representation, decides the next action, and the runtime executes it. They unlock the long tail of systems without APIs, at the cost of speed, flakiness on layout changes, and detection by anti-bot systems.
What is an example of Browser Agent?
An agent logs into a supplier portal weekly, downloads the invoice PDFs and files them in Drive — a portal with no API and no export.
Why does Browser Agent matter for AI and automation?
A browser agent drives a real web browser — navigating, clicking, typing and reading pages — to complete tasks on sites that have no API. It connects to the workflows, prompts and tool stacks linked on this page, so you can move from definition to execution without leaving Onexial.