Web scraping without writing the scraper
Orchestra is a desktop app for pulling structured data off websites. You point at what you want, it works out the selector, and you get JSON or CSV — plus the plain Playwright code underneath, so you are never locked in.
What makes web scraping hard is rarely the scraping
Fetching a page is easy. What costs the afternoon is everything around it: the site renders its table in JavaScript so the HTML you downloaded is empty, the results are paginated behind a button that only appears after scroll, the data you need is behind a login, and the selector you wrote breaks the following Tuesday because a class name changed.
Orchestra runs a real browser, so JavaScript-rendered pages are simply pages. You record against the live site, and the steps you record are the steps that run.
How extraction works
A scrape in Orchestra is a flow: a short list of visual steps you can read top to bottom.
- Point and pick
- Click an element in the live preview. Orchestra proposes a selector, shows you what it matches, and lets you loosen or tighten it before you commit.
- Extract List
- Select one row and it generalises to all of them — products, listings, table rows, search results — with each field named the way you want it in the output.
- Each and pagination
- Loop over rows, follow every detail link, or keep clicking "next" until the button stops existing. Nesting works, so a list of categories containing a list of products is two Each steps.
- Output
- Write JSON or CSV wherever you want it. The shape is yours to name, not a dump of the DOM.
Logins, sessions and the things that block scrapers
Sites that need an account are the normal case, not the exception. Orchestra drives a real browser profile, so you sign in once the way a person would and the session persists across runs. There is no cookie-jar surgery and no reverse-engineering an auth flow.
Because it is your machine and your browser, the traffic looks like what it is. That also means the responsibility is yours: read the terms of the site you are scraping, respect robots.txt and rate limits, and do not collect personal data you have no basis to hold.
You keep the code
Every flow exports to plain Playwright — a readable TypeScript file with no Orchestra runtime, no SDK and no license check. Drop it in a repo, run it in CI, hand it to a developer, schedule it with cron.
That is the difference between a tool and a dependency. If Orchestra vanished tomorrow, your scrapers would keep running.
Common questions
Do I need to know how to code to scrape a website with Orchestra?
No. You build the flow visually by pointing at elements in a live browser preview. Code is the output, not the input — you can read and export the Playwright it generates, but you never have to write it.
Can it scrape JavaScript-heavy sites and single-page apps?
Yes. Orchestra drives a real browser, so anything a browser can render is scrapable — React, Vue, infinite scroll, content that only appears after an XHR completes. There is no separate "JS rendering" mode to enable.
What formats can I export the scraped data to?
JSON and CSV, with the field names and structure you define in the flow. You can also export the flow itself as plain Playwright code and write the output anywhere your own script chooses.
Does the scraper run in the cloud?
No. Orchestra is a desktop application and runs entirely on your own machine, on Windows, macOS or Linux. Nothing you scrape passes through our servers, because there are no servers in the path.
How much does it cost?
There is a 14-day free trial with every feature unlocked and no card required. After that it is $149 once — a one-time purchase, not a subscription, with no per-run or per-page charges.