How to Start Web Scraping With Bright Data Proxies

November 18, 2026

Bright Data is a proxy and web-data platform with roughly 150 million residential IPs, a scraping browser, and prebuilt datasets you buy instead of crawl. Monthly pricing is commitment-based; the proxy network starts around $5.88 per GB of residential traffic.

What is Bright Data?

Bright Data is an Israeli infrastructure company, founded in 2014 as Luminati and renamed in 2021. It sells three things that are often confused with each other:

  • Proxy network — residential, datacenter, ISP and mobile IPs, billed by bandwidth or per IP.
  • Web Scraper API and Scraping Browser — an endpoint that handles proxy rotation, CAPTCHA solving and JS rendering for you, returning HTML or parsed JSON.
  • Datasets — prebuilt, refreshed collections covering Amazon products, LinkedIn profiles, Instagram, Indeed jobs and dozens more, sold per record.

Most buyers arrive expecting to rent proxies and discover they actually need the Scraper API, because raw proxies still leave you to write the scraper. Knowing which one you need before signing up saves real money.

How do you get started?

The onboarding is heavier than a typical SaaS tool, deliberately — they are selling to data teams.

  1. Sign up and complete business verification. Bright Data requires KYC for residential proxy access; individual hobbyist accounts get a limited sandbox.
  2. Open the dashboard and pick a product. The proxy zone manager is where you create a “zone” — a named pool of IPs with a country and session policy. Zones are the unit everything bills against.
  3. For scraping, add the Web Unlocker zone instead. It takes a target URL and returns rendered HTML with rotation handled.
  4. Grab your credentials: host, port, username and password, or the API token for the REST endpoints.
  5. Test with a single request before building. curl -x http://user:pass@brd.superproxy.io:33335 https://geo.brdtest.com/mygeo.json returns your apparent geography — the fastest way to confirm a zone works.

Do not skip step 1. Accounts stuck in verification are the most common “it does not work” complaint in their developer forums.

What does it cost in 2026?

Product Price Notes
Residential proxies From about $5.88/GB Volume tiers drop the rate
Datacenter proxies From about $0.60/IP Fast, easily blocked
ISP proxies Higher per IP Static, trusted reputation
Web Unlocker About $1.50 per 1,000 requests Handles CAPTCHA and JS
Scraping Browser About $8 per GB of traffic Full browser sessions
Datasets Per record, varies Amazon product around $0.001-0.01 each

The pricing structure is where people get burned. Residential bandwidth is metered per gigabyte, and a JS-heavy page with images can pull 3-8 MB. Simple math: 10,000 product pages at 5 MB each is 50 GB, or roughly $290, before you have written a single line of parsing logic.

Turning off image loading in the Scraping Browser cuts that dramatically. Do it on the first request, not the hundredth.

How does the Scraper API handle the hard parts?

The value proposition is that three failure modes stop being your problem:

  • CAPTCHAs: the unlocker retries with different fingerprints and solves where it can.
  • JS rendering: headless browser runs server-side, so you get the DOM after hydration.
  • IP blocking: rotation happens per request within your zone’s rules.

What it does not do is tell you the target site changed its layout. Your parser still breaks when Amazon moves a div, and that is on you to notice. Budget monitoring; a silent 40% parse-failure rate on a paid dataset is worse than a crash.

One practical tip: use session_id in your proxy username when you need the same IP across requests — for a login flow or a paginated scroll. Rotating mid-flow logs you out and looks like a bot.

Where does Bright Data fall short?

  • Cost scales with complexity. JS-heavy sites with images can hit hundreds of dollars a month even at moderate volume.
  • Verification friction. Business KYC for residential access excludes hobbyists and slows down small teams.
  • Support quality varies by plan. Low-tier accounts get community support; enterprise gets an account manager.
  • Compliance responsibility sits with you. They require lawful use, but the GDPR and CFAA exposure for what you scrape is your own to manage. Do not scrape personal data casually.
  • Dashboard is dense. The number of products and settings makes it easy to configure a zone and misread what it will bill.
  • Datasets are snapshots with refresh schedules, not live queries, so “real-time pricing” needs the proxy path instead.
  • Documentation assumes you know what a zone is — the chicken-and-egg is real for first-timers.

Who should use Bright Data?

Data engineering teams building or maintaining large-scale collection at millions of pages monthly, price-comparison and market-intelligence products, and teams that would rather buy a maintained dataset than build one. Enterprise compliance and legal departments usually prefer their structure to a smaller reseller.

Skip it if you need a few hundred pages once — a simple script plus a cheap residential provider costs less than a month of this. Skip it if you have no engineering capacity, since nothing here is point-and-click for custom targets. And skip it if the data you want is personal information and you have not done the legal analysis; this tool will happily help you violate GDPR.

Quick Summary

  • Bright Data (founded 2014 as Luminati) sells residential proxies, a scraping browser, and prebuilt datasets.
  • Residential traffic starts around $5.88/GB; Web Unlocker is about $1.50 per 1,000 requests.
  • Image-heavy scraping can reach hundreds of dollars monthly — disable image loading early.
  • KYC verification is required for residential access, which blocks hobbyist accounts.
  • It handles CAPTCHA and rotation, but schema changes at the target remain your problem.