Brightdata

Overview

Bright Data is a web data infrastructure company: proxy networks, unblocking/scraping APIs, a browser API, pre-collected datasets and an MCP server for AI agents. It is headquartered in Tel Aviv, Israel; founded in 2014 as Luminati Networks (originally tied to Hola VPN), bought by London fund EMK Capital in 2017 (reported about USD 200M) and renamed Bright Data in March 2021 (Wikipedia). CEO: Or Lenchner (Wikipedia). Its about page claims 50,000+ organisations as customers, 14 of the top 20 LLM labs and ARR above USD 300M as of 2025 (vendor claims, unverified).

Legal: Meta (2023) and X Corp (2023) sued over scraping; US federal judges ruled for Bright Data in January 2024 (Meta) and May 2024 (X Corp), finding that scraping publicly accessible data did not breach the terms/copyright claims at issue (Wikipedia).

Products and plans (brightdata.com pricing pages, 2026-10-02)

ProductWhatBilling
Web Unlocker and SERP APIUnblocking requests; multi-engine search resultsPay-as-you-go per 1,000 requests (“pay only for success”); monthly Scale plan with a discounted overage rate; free tier of 5,000 requests/month (checked 2026-09-30)
Web Scraper APIPre-built scrapers for popular sites, JSON/CSV outputPay-as-you-go per 1,000 records; Scale plan; free 5,000 records/month
Browser API (Scraping Browser)Remote stealth browsers for Puppeteer, Playwright, SeleniumPay-as-you-go per GB; Scale plan with included traffic; free 5,000 credits/month incl. 1 GB
Residential proxiesRotating residential IPs, geo-targeting to zip levelPay-as-you-go per GB (promotional discounts shown); monthly plans with included traffic; site claims 400M+ IPs in 195 countries
Datasets marketplacePre-collected, refreshed dataOne-time per-record purchase; subscription discounts for refresh (monthly 80%, quarterly 50%, biannual 25%); 120+ domains, 190+ datasets listed
Scraper Studio, Crawl API, Web Archive, ISP/datacenter/mobile proxies, Retail Insights, managed acquisitionNot re-verified (see Open items)

Promotions (discount codes, deposit match) change often. Current prices: see the vendor pricing pages https://brightdata.com/pricing

AI and agent support

  • MCP server: brightdata/brightdata-mcp (MIT, 69 tools at last check 2026-09-30; hosted or npx @brightdata/mcp; 5,000 free requests/month). Vendor states it works with Claude, Cursor and other MCP clients.
  • Agent-oriented products listed on the site include Search and Extract and agent browser features (vendor descriptions).
  • For where its SERP API fits among search APIs see ai-search-apis-overview.

Strengths and limitations

  • Strengths (opinion, drawn from vendor material): broad product range, success-based pricing on some products, an MCP server, court rulings cited above.
  • Limitations (opinion): usage-based pricing across many products may be complicated; the ethics and legality of scraping remain contested (see litigation above).
  • Comparison to Supadata (video/web transcript extraction API) and ScrapingBee-type tools: Bright Data is the broad infrastructure vendor; the earlier note’s “better pricing” claim was unsupported and is dropped.

Open items

  • Unverified: pricing for ISP, datacenter and mobile proxies, Scraper Studio, Crawl API and Web Archive; Retail Insights and managed-service prices; the older “1,000-page-load free trial” (the site now shows free tiers instead).
  • Unverified vendor claims removed from this version: 99.99% uptime, under-10-minute support response, Trustpilot/G2/Capterra ratings, 20,000+ customers (the about page now says 50,000+), ISO 27001 certification (the about page does not mention it; the trust-center page returned 404), Bright Initiative partner counts beyond Wikipedia’s 700+ organisations and 300+ academic institutions.
  • Residential IP count: old note 150M+, site now 400M+ (vendor claim).
  • No independent reviews were consulted.

Sources (accessed 2026-10-02 unless noted)