TestMu AI Browser Cloud

TestMu AI Browser Cloud is a coding tool. A cloud browser service from TestMu AI that gives agents and automation scripts isolated real Chrome sessions, with session persistence, best-effort stealth, a localhost tunnel and automatic session replay and logs.

TestMu AI Browser Cloud supplies AI agents and automation scripts with real Chrome browsers running in the cloud. Each session is an isolated Chrome instance, so there is no machine to provision, no headless binary to install and no warm pool to look after. Sessions can be started on demand, from a single one up to hundreds in parallel, with the available concurrency depending on the chosen plan. The product is aimed at developers building browser-driving agents, QA engineers moving existing suites off local infrastructure, and teams that need rendered web pages at scale. It is installed as an npm package (@testmuai/browser-cloud). A script creates a session through the SDK and then connects to it using Puppeteer, Playwright or Selenium, or through the agent's own code. According to the vendor, existing scripts usually need only minor changes. Cookbooks cover agents built on OpenAI, Claude and Gemini computer use, plus plain Playwright and Puppeteer setups. Several features target agent work specifically. Session persistence keeps cookies, local storage and login state between runs, which avoids repeated authentication. A best-effort stealth mode masks fingerprints and blocks ads to reduce the number of blocked requests, and the vendor is careful to call it best-effort rather than a guarantee. A built-in localhost tunnel lets cloud browsers reach internal apps, localhost and VPN-gated services without a third-party tunneling tool. Every session automatically records video, network logs, console output and step-by-step command replay, which makes failed runs traceable. Typical uses listed include QA against real browsers, scraping JavaScript-heavy pages (within each site's terms and applicable laws), multi-step workflows that keep client-side state, price and inventory monitoring, and collecting rendered content for model training. The infrastructure is the same cloud behind the vendor's cross-browser testing grid, so it comes from an established testing platform rather than a new agent-only startup. Compared with agent-focused browser services, the testing heritage shows in the debugging tools and CI/CD integrations. The Chrome-only agent focus and plan-based concurrency are points to check before committing.