---
title: "Web and API testing in a browser"
description: >-
  Learn how to build browser-based Web tests and API steps, configure Web
  agents, and run automated checks against your sites and services.
sidebarTitle: "Web"
lastUpdated: "2026-09-23"
---

> **For AI agents:** the complete documentation index is at [llms.txt](/llms.txt). Append `.md` to any page URL for its markdown version.

The Web channel runs browser-based tests against websites and web apps. API steps hit HTTP services directly, without a browser. Both run on the **Web agent**, a machine or server with network access at a specific location.

![The web test builder: a Navigate to starting page action followed by a Fill Form step group with hover, click, and fill-input steps, plus a Preview/Run panel and recent executions](/images/web-test-builder.webp)

## Quickstart

Create your first browser test in four steps.

<Steps>
  <Step>
    **Create the test.** Open the **Tests** section for your product, click **+ Create new test**, and choose **Web (New experience)**. Fill in the test name, the default Web agent, and the start URL the browser opens.
  </Step>
  <Step>
    **Add steps.** In the builder, record interactions on the live page or add web actions manually: click an element, fill an input, wait for content, and assert that something is displayed. Save the test.
  </Step>
  <Step>
    **Run it on a browser.** Click **Run** in the builder to watch it live, or **Run** on the test's row, then confirm the agent and variables on the [run confirmation](/executions/run) page and launch.
  </Step>
  <Step>
    **Read the result.** Open the execution under **Executions** to see status, per-step screenshots, and timing.
  </Step>
</Steps>

<Tip>
The browser type, window size, language, and user agent come from the [browser preset](/tests/browser-presets) you pick in the builder's **Environment** tab.
</Tip>

## Web tests

Web tests automate user interactions like clicking, typing, navigating, and checking page elements.

**Starting point:** you provide a URL template. Tests launch the browser, navigate to that URL, and run your steps from there.

**Browser options you can configure:**
- Browser type
- Custom user agent
- Language/locale
- Window size (default 1920×1080)

**Common actions:** click elements, type text, navigate, wait for elements, check that elements exist or contain text, scroll, hover, and read element properties.

## API steps

Test REST and HTTP services directly with [API steps](/tests/actions/api/overview) added inside a web or smartphone test, useful for backend validation, contract checks, and end-to-end flows that mix UI and direct backend calls.

**Request configuration:**
- HTTP methods: GET, POST, PUT, PATCH, DELETE
- URL with templated variables
- Custom headers
- Body: JSON or raw payload
- DNS override, proxy routing, SSL handling, redirects, and cookies

**Response validation:**
- Status code checks
- Body and header assertions
- Variable extraction from the response for use in later steps

**Per-request metrics:** detailed timing breakdown (DNS lookup, TCP connection, TLS handshake, content generation, transfer, total) for performance analysis.

## The Web agent

A server or machine that runs browser-based tests from a specific network location.

- Automates website and web application testing
- Launches browsers and executes test steps
- Supports multiple simultaneous tests

**Configuration:** hostname or IP address, parallel test capacity, automatic reboot and update scheduling, network configuration, and credentials for authentication.

Web agents report their own operational states: see [Statuses](/equipment/statuses), including the setup-only "Waiting for Config" state unique to web agents.