Browser Testing Launched

Until now, LoadForge specialised in high-scale HTTP load tests—perfect for pushing millions of virtual users against your APIs. That’s still core, but some questions live inside the browser: Largest Contentful Paint (LCP), flaky JavaScript flows, third-party widgets, visual shifts... Today you can test those too.

Why we built it

1. True user-level performance – surface Core Web Vitals (LCP, CLS, FID) and Lighthouse metrics that request-level tests miss.
2. End-to-end functional coverage – click buttons, fill forms, follow journeys; catch regressions before users do.
3. Production monitoring – schedule low-frequency browser checks (5–10 sessions per generator) and watch metrics drift before a deploy tanks them.
4. Unified workflow – stay inside LoadForge; no extra tools or agents needed.

Under the hood

Each LoadForge generator now embeds Playwright to run headless Chrome. You script tests with familiar Locust syntax using the PlaywrightUser class:

from locust_plugins.users.playwright import PlaywrightUser, PageWithRetry, pw, event
from locust import task

class JourneyUser(PlaywrightUser):
    @task
    @pw
    async def journey(self, page: PageWithRetry):
        async with event(self, "Home"):
            await page.goto("/")
        async with event(self, "Checkout"):
            await page.click('button:has-text("Buy now")')

Events flow back to your LoadForge dashboard just like classic Locust requests—so graphs, percentiles, and pass/fail counts “just work”.

Perfect-fit use cases

• Core Web Vitals regression guardrails – fail if LCP > 1.2 s or CLS > 0.1.
• Critical user-journey validation – “search → product → cart → checkout” every 15 minutes.
• Visual smoke tests – capture screenshots after deploy and detect drift.
• Accessibility scanning – inject Axe and raise an error on violations.
• Edge-network performance – throttle to 3 G to ensure first paint < 2 s.

Need inspiration? Check the six ready-made scripts in the docs.

Know the limits

Browser sessions are heavier than raw HTTP requests. A generator that drives tens of thousands of virtual users for load tests will handle ~5–10 simultaneous Chrome sessions. Browser testing is for UX & quality, not for maximum-TPS benchmarking. Combine it with classic load tests for full coverage.

Get started now

1. Open LoadForge and select Browser Tests from the left-hand menu.
2. Create a new run using one of the example scripts or your own Playwright script.
3. Launch and watch real Chrome sessions execute live, with metrics flowing into familiar LoadForge graphs.

Browser testing is live on every paid plan today. Spin up a run, watch Chrome fly, and unlock the UX layer of your performance story.