One API to automate any browser workflow

Give it a URL, a task in plain English, and a schema. Get back structured JSON. No Playwright scripts. No selectors. No maintenance.

example.py
from computeruse import ComputerUse

cu = ComputerUse()

result = cu.run_task(
    url="https://news.ycombinator.com",
    task="Get the top 5 posts with title, points, and link",
    output_schema={"posts": [{"title": "str", "points": "int", "link": "str"}]}
)

# => {"posts": [{"title": "Show HN: ...", "points": 342, "link": "..."}]}

How it works

1

Describe your task

Pass a URL, a plain-English instruction, and an optional output schema. No selectors or scripts needed.

2

Agent executes

An AI agent launches a real browser, navigates the page, fills forms, clicks buttons, and extracts data.

3

Get structured data

Receive clean JSON matching your schema. Every run is recorded with step-by-step screenshots.

Simple, step-based pricing

One step = one browser action. Pay only for what you use.

Free

$0forever

500 steps/month

  • 1 concurrent task
  • 50 steps per task
  • Community support
Get Started Free
Popular

Startup

$29/month

5,000 steps/month

  • 5 concurrent tasks
  • 200 steps per task
  • 5 min timeout
  • Email support
Get Started

Growth

$99/month

25,000 steps/month

  • 10 concurrent tasks
  • 350 steps per task
  • 7.5 min timeout
  • Priority support
Get Started

Enterprise

Custom

100,000+ steps/month

  • 20+ concurrent tasks
  • 500 steps per task
  • 10 min timeout
  • Dedicated support
  • SLA guarantee
Contact Us