Pricing Docs MCP

Human-in-the-Loop
QA Testing

Call our API, get real human testers to verify your app, receive AI-extracted structured results. Perfect for CI/CD pipelines, pre-deployment checks, and bug verification.

1const job = await runhuman.createJob({
2  url: "https://myapp.com/checkout",
3  description: "Test checkout flow",
4  schema: {
5    canCheckout: { type: "boolean" },
6    uxIssues: { type: "string" }
7  }
8});
9
10// Wait for human tester result
11const result = await job.waitForResult();
12// { canCheckout: true }
<2min average response time Video recordings 99.9% uptime

Secure & Private

API key authentication, HTTPS encryption, and optional private tester pools. Your test data stays secure.

Secure & Private

CI/CD Integration

Perfect for GitHub Actions, Jenkins, or any CI/CD pipeline. Add human verification to your automated deployment workflow.

CI/CD Integration

Flexible Schemas

Use our templates (login-flow, form-submission, etc.) or define custom schemas to extract exactly the data you need from testers.

Flexible Schemas

Real Testers Feedback

Your product is tested by real humans under real conditions. Catch UX gaps, logic flaws, and overlooked details before they reach your users.

Real Testers Feedback
Pricing

10x Cheaper Than AI Computer Use

On-demand usage billing at $0.0018/sec of human testing time. No monthly fees, no minimums, no commitments.

20-step test:
$0.189
humans
vs
$1.94
Opus 4.5
Costs scale linearly, not exponentially
No token costs for screenshot processing
POPULAR

General Use

Most popular for CI/CD pipelines

$0.0018 /second
Wait Time
<5 minutes
SLA Guarantee
Full refund if missed
Example Cost
60s test = $0.108
120s test = $0.216
✓ Real human testers
✓ Video recordings included
✓ AI-extracted results
✓ API & MCP support

Use Runhuman with
any framework

How it works

Start testing in seconds

Three steps to get human-verified QA results

Call the API

Call the API

Make a POST request with your URL, test instructions, and desired result schema. Your test is queued instantly.

Human Tests It

Human Tests It

A real human tester performs your test and describes the results in natural language.

Get Structured Results

Receive structured JSON data matching your schema, ready for your pipeline.

Trusted by developers worldwide

@alexdev @alexdev

I build AI workflows that need human taste and judgment at certain steps. Before RunHuman, that meant stopping everything and doing it myself, or managing testers. But now, i have access to people via API, a Task Rabbit for Claude to use to get unstuck or to verify he's not building something insane. It's incredible that my AI agents can run humans without me!

@mikejy @mikejy

Integrating Runhuman was a game-changer for my AI agents. Real human feedback in the loop means they actually get things right the first time.

@sarahk @sarahk

I integrated Runhuman into all our CI pipelines. It's very cheap and fast using the Instant tier for critical flows and the General Use tier for everything else. We're shipping faster than ever with real people validating our features, UX, and overall experience on every push. QA as a Service!

@davidchen @davidchen

We trust our releases more now. That confidence alone is worth it.

@CarlosRivera @CarlosRivera

RunHuman helps me catch the things automated tests miss. With a simple API call, I get real human feedback that makes my releases more reliable and my AI workflows feel safer.

@kenji_m @kenji_m

Getting real human validation in my development workflow has been incredible. The structured feedback helps me iterate faster and ship with confidence.

@tyler-mills @tyler-mills

Simple API, real humans, actual results. RunHuman gives me the human-in-the-loop validation I need without the overhead of managing testers.

@priya

I signed up and went through the full onboarding flow. Account creation was very smooth and fast. I tested the Playground and explored how tests are set up in UI mode (URL selection, test instructions, timing, and output schema). I tried running a real test on a live website (Zara DE) to understand the full flow. At that point, I realized that running tests requires creating an API key, which makes sense since the product is clearly built for real, production-level use rather than a simple demo. Even without running the final test, the experience made it very clear how Run Human is intended to be used: it's developer-focused, structured, and designed for serious human-in-the-loop QA rather than a toy tool. The boundaries between setup, execution, and human validation are very nice.

Ready to add human QA to your pipeline?

Get started in minutes with our simple API.