Load test the flows your users really take.

Any API flow you record can run as a load test. No second script to write or keep in sync.

Create a free workspace Read the load test docs

A load test running at 3:32 of 4:30: passing so far with p95 26 ms and 0.02% errors at 400 virtual users and 748 requests per second. A chart shows requests per second rising with virtual users while p95 stays flat, well under the 400 ms limit. — A load test in Stresseur Cloud, live: virtual users, req/s, p95 and errors against the profile's thresholds. Sample data.

One flow file. Two jobs.

On every pull request

The flow runs as a test. It checks each response and shows which step got slower than on main.

Under load

The same flow runs as many virtual users at once, with the login token and each new ID still passed between steps. Change the flow once, and both jobs follow.

A load profile is a few lines of YAML.

It lives next to your flows in api-tests/stresseur.load.yaml. Without one, you get Smoke, Load, Stress, Spike, Soak and Breakpoint profiles.

api-tests/stresseur.load.yamlexcerpt

profiles:  - name: load    flows: { storefront-browse: 70, admin-catalog-crud: 30 }1    executor: ramping-vus2    stages:      - { duration: 1m, target: 200 }      - { duration: 3m, target: 200 }      - { duration: 30s, target: 0 }    think_time: { min: 1s, max: 3s }    thresholds:3      p95: 300ms      errors: 1%      steps: { Checkout: { p95: 500ms } }    abort: { errors: 20%, window: 30s }4

1
Your real flows, mixed
The same YAML flows your pull requests run, weighted like real traffic.
2
Users or arrival rate
Constant or ramping virtual users, or a constant or ramping request rate.
3
Pass or fail on numbers
p95 and error-rate thresholds for the whole test, or for one step.
4
Stop before it hurts
An abort rule ends the run when errors or p95 stay too high.

Watch it live.

Virtual users, req/s, p95, errors and the latency of each step stream while the test runs.

See where it breaks.

The chart marks the point where p95 first crossed your limit, with the users and req/s at that moment.

Load only what you own.

Load goes only to domains your workspace has verified. Secrets stay encrypted in its environments.

From 4 regions.

Frankfurt, Virginia, San Jose and Singapore. Load generators start in the regions you pick.

Need more for a launch? The $499 Launch pack gives a 2-hour window of tests up to 5,000 virtual users.

Plan	VUs	Length	Regions	VUh/mo
Free	50	10 min	1	50
Team	1,000	1 hour	All 4	5,000

VUh means virtual-user hours. See pricing

When a script is the better call.

Tools like k6 give you full control in code: custom protocols, unusual pacing, load shapes no user would make. If you need that, write the script. Stresseur is for load testing the journeys your users take, without a second copy of them.

Or start on your laptop.

The open-source CLI runs a flow under load locally, for free.

$npx stresseur flow run api-tests/checkout.yamlflow.yaml checkout --vus 10 --duration 60s

Find the breaking point before your users do.