Term
Browser Use
Browser use refers to AI agents that drive a real web browser — navigating, clicking, filling forms, reading content — to complete web tasks autonomously. The browser serves as both the tool and the perception surface.
Browser Use — explained in detail
Browser use describes the pattern in which an AI agent uses a web browser as its execution environment to solve multi-step tasks on the web. The agent receives the page state — as a screenshot, as a cleaned-up DOM tree (often the accessibility tree) or both — and returns actions: open a URL, click an element, type text, scroll or extract the visible content. A control layer such as Playwright or Puppeteer performs these actions in a real browser instance and returns the new page state.
The term also gained traction through the open-source project of the same name, Browser Use, which gives LLMs a standardized interface for controlling a browser. Compared to purely DOM-based scripts, the approach adds robustness because the agent understands semantically what needs to be done on the page instead of relying on brittle CSS selectors.
Example / Practical use
Typical tasks: find and compare flight options on a booking portal, collect data from several pages behind a login, fill in a web form with order details, or run an end-to-end test of a web app. In practice browser use is slower and more error-prone than a direct API integration: every action requires an inference step, dynamic pages and cookie banners disrupt the flow, and a single wrong click can derail the sequence. Wherever an API is available, it remains the preferable option.
Distinction from related terms
Computer use is the more general case: there the agent operates an entire operating system including arbitrary desktop applications, whereas browser use is limited to the browser. Classic browser automation with Playwright or Selenium runs through deterministic, pre-written scripts without any model decision — browser use leaves the step planning to the AI agent. It differs from function calling in that the tool is not a defined function but an open web interface.
Entdecke mehr
AI Workflows by Keyword: How We Make Recurring Routines Enforceable
A typed keyword triggers a fixed AI routine — and every single step must be committed before the next one appears. Why that's the actual trick.
GlossarEnsemble / Multi-Model Orchestration
Ensemble means combining several deliberately varied LLM runs or models whose findings complement each other. Multi-model orchestration drives these runs via orchestrators with sub-agents, so the union of results is larger than any single run.
LexikonFunction Calling / Tool Use
How an LLM uses tools: define a tool as a schema, the model picks the function and arguments, the result returns to the chat — the basis of every agent.