
Bright Data
Bright Data is a web data platform combining a 400M+ IP proxy network, ready-made scraper APIs, pre-collected datasets and agent-ready web access, so technical teams can gather public web data without building the collection infrastructure themselves.
What is Bright Data?
Bright Data sells the plumbing behind large-scale public web data collection rather than a single application, and the catalogue is organised into four blocks. Web Access APIs handle the hard part of reaching a page: the Unlocker API works around blocks and CAPTCHAs, the SERP API returns structured results from Google, Bing, DuckDuckGo and Yandex, the Browser API runs remote auto-scaling browsers compatible with Puppeteer, Selenium and Playwright, and the Crawl API turns whole sites into model-ready data. Data Feeds cover delivery: Scraper APIs for more than 800 sites, Scraper Studio for the ones they do not cover, a Datasets marketplace drawing on over 600 domains, a Data Firehose that streams records as they are collected, and a petabyte-scale web archive the company says holds 429 billion cached pages. Data and Insights adds Bright Insights for e-commerce price and catalogue intelligence, a Managed Data Acquisition service and the Deep Lookup beta. Underneath sits the proxy infrastructure: more than 400 million residential IPs across 195 countries, 1.3 million ISP IPs and 1.3 million datacentre IPs, with targeting down to city, ZIP code, carrier and ASN, and advertised figures of 99.99% network uptime and a 99.95% success rate. A separate Data for AI branch aims at model builders and agents: multimodal video and audio training data, LLM-ready data packages, Search and Extract, an Agent Browser, and a Web MCP server that hands an assistant search, crawl, access and navigate tools with 5,000 free requests a month. Compliance is presented as a differentiator: an opt-in peer network, a Know Your Customer review before certain services open, a published Acceptable Use Policy, monitoring of more than 30 billion domains and a multilingual Compliance and Ethics team, backed by ISO 27001, 27017, 27018 and 27701 certificates, SOC 2 Type II and SOC 3 reports, a CSA STAR listing and an independent PwC assurance report. Founded in 2014 as Luminati, renamed Bright Data in 2021 and majority-owned by EMK Capital since 2017, it claims more than 50,000 customer organisations and over 5,500 granted patent claims.
What it does
- Collect structured data from more than 800 websites through ready-made Scraper APIs
- Buy pre-collected datasets covering over 600 domains, from LinkedIn and e-commerce to social media and real estate
- Route traffic through 400M+ residential, 1.3M ISP and 1.3M datacentre IPs, targeting country, city, ZIP code, carrier and ASN
- Get past blocks and CAPTCHAs automatically with the Unlocker API, and render dynamic pages with the Browser API
- Pull live search engine results from Google, Bing, DuckDuckGo and Yandex through the SERP API
- Give an AI agent live web access - search, crawl, access, navigate - through the Web MCP server
- Build a scraper for any site in Scraper Studio, or hand the whole job to the Managed Data Acquisition team
When to use Bright Data / When not to
A quick filter to help you decide if Bright Data is the right fit.
When to use Bright Data
- Engineering and data teams that need public web data at scale and would rather not build and maintain proxies, unblocking and parsers themselves
- AI teams sourcing training, fine-tuning and RAG material, or giving an agent live web access through the Web MCP server and the Agent Browser
- E-commerce and retail organisations tracking competitor prices, catalogues, availability and reviews through Scraper APIs, Datasets and Bright Insights
- Market research, SEO and marketing teams that need SERP rankings, social media data and audience research collected continuously rather than by hand
- Buyers who want data delivered rather than collected: the Datasets marketplace, the no-code scrapers and the Managed Data Acquisition service require no engineering at all
When not to use Bright Data
- Individuals and hobby projects: the Master Service Agreement is a business contract, several services only open after a Know Your Customer review that can include video calls, and the minimum age is 18
- Occasional or very small workloads: the free tier stops at 5,000 records a month and the first committed monthly tier starts at USD 499
- Competitors: the agreement forbids using the service to build or run anything that competes with Bright Data, and datasets may not be redistributed as a similar product
- Resellers, who need Bright Data's prior written authorisation before passing the service on in whole or in part
- Organisations that need a named hosting country, a published subprocessor list or a documented opt-out from the vendor's reuse of collected data: none of the three is published
How to use Bright Data
A typical end-to-end flow, from setup to results.
- Create an account from the site with Start free trial or Get started for free; no credit card is required
- Clear the compliance review: some services only open after a Know Your Customer process that can include video calls
- Pick the right block: Datasets if the data already exists, a Scraper API if the site is covered, Scraper Studio if it is not, raw proxies if you run your own scraper
- Collect your credentials in the dashboard and follow the documentation on docs.brightdata.com
- Integrate through the REST API, the Python SDK, the JavaScript SDK or the command-line interface
- For browser-driven work, point Puppeteer, Selenium or Playwright at the Browser API
- For an AI agent, install the Web MCP server from the public GitHub repository and connect it to your assistant
- Choose how data arrives: webhook or API, parsed to JSON or CSV
- Set monthly spend limits if you stay on pay-as-you-go
- Install the browser extension for one-click proxy access from any major browser
Pros & Cons
Pros
- Unusual breadth for a single account: proxies, scraping APIs, datasets and agent tooling on one platform
- One of the largest advertised proxy pools on the market, with more than 400 million residential IPs across 195 countries
- Success-based billing on several products: nothing is charged for a failed delivery
- Permanent free tiers with no credit card, including 5,000 records a month on the Scraper API and 5,000 Web MCP requests
- Independent certifications: ISO 27001:2022, 27017, 27018 and 27701, SOC 2 Type II, SOC 3, CSA STAR and a PwC assurance report
- Article 27 representatives designated for both the EU (EDPO, Brussels) and the UK (EDPO UK, London), plus a published Data Protection Addendum
- Round-the-clock support with an advertised sub-10-minute average response time and a standard SLA on every plan
Cons
- Pricing is scattered across one page per product, each with its own unit - per 1,000 requests, per 1,000 records, per GB, per IP or per month - which makes the real cost hard to anticipate
- The entry ticket jumps once you leave pay-as-you-go: USD 499 a month for the first committed tier, a USD 500 monthly minimum for Premium support and USD 10,000 a month for Enterprise support
- No hosting country and no region are named anywhere; the Trust Center says only AWS with multiple Availability Zones
- No named subprocessor list is published, only categories of recipients such as cloud services and affiliated companies
- The agreement lets Bright Data retain and reuse the data a customer has collected at its sole discretion, and no opt-out is documented
- The CCPA notice acknowledges that identifiers may have been sold to third parties over the previous twelve months
- Access is gated by a compliance review that can require video calls and identity documents, and no retention period is ever quantified
Pricing & Plans
A permanent free tier is available without a credit card, capped at 5,000 records per month on the Web Scraper API and 5,000 requests per month on the Web MCP server. Paid usage begins on a pay-as-you-go basis, from USD 0.20 per 1,000 archived HTML pages, USD 1.00 per 1,000 requests on the Unlocker, Crawl and SERP APIs, USD 1.50 per 1,000 records on the Web Scraper API and USD 8.00 per GB on residential proxies. The lowest committed monthly plan is priced at USD 499 per month. All charges are contractually billed in USD; amounts displayed in local currency are indicative only.
- 5
- 000 records per month
- no credit card
- expert support
- USD 1.50 per 1
- 000 records
- monthly spend limits
- unlimited concurrency
- USD 499 per month
- 384
- 000 records included
- USD 1.30 per 1
- 000 additional records
- cancel anytime
- custom pricing
- volume discounts
- account manager
- premium SLA
- priority support
- SSO
- USD 8 per GB
- no monthly commitment
- USD 499 billed monthly
- USD 999 billed monthly
- USD 1
- 999 billed monthly
- custom pricing
- unlimited seats
- premium SLA
- 99.99% network uptime
- free Proxy Manager
- included with every plan
- business days
- standard SLA
- 24/7
- 8% of the monthly invoice up to USD 5
- 000 per month
- minimum USD 500 per month
- 24/7
- USD 10
- 000 per month
- dedicated technical account manager
- Slack channel
- onboarding specialist
Data, GDPR & hosting
A consolidated view of how Bright Data handles your data.
GDPR overview
GDPR and CCPA compliance is claimed explicitly: the Trust Center states Bright Data is fully committed to complying with both, and the privacy policy, last reviewed on 14 May 2026, lists the seven EU rights (confirmation and access, portability, rectification, erasure, objection, restriction, complaint to a supervisory authority), exercised in writing at privacy@brightdata.com where a Privacy Officer investigates. The company is established in Israel and limits transfers outside the EEA to countries the European Commission recognises as adequate or to recipients covered by contractual safeguards. Article 27 representatives are named for both jurisdictions: EDPO in Brussels for the EU, EDPO UK in London for the UK. A Data Protection Addendum is published, and the security posture is backed by ISO 27001, 27017, 27018 and 27701, SOC 2 Type II, SOC 3, CSA STAR and a PwC assurance report.
Who owns the data?
The agreement draws a sharp line. Bright Data states it does not rent or sell User Data, and it never discloses one customer's IP address to another. For the data a customer collects through the service, however, the Master Service Agreement reserves broad rights: Bright Data may retain that data and use it for its own purposes at its sole discretion, and the same wording covers the Web Scraper IDE. Delivered datasets and exports are kept only for a limited Review Period, after which they may be deleted with no obligation to store or re-deliver them. Where personal data is processed, Bright Data acts as processor under a published Data Protection Addendum and the customer remains the controller.
Reuse rights
Customers may use the data they collect and the datasets they buy for their own lawful business purposes without asking permission each time, subject to three limits. Datasets may only serve legally valid purposes and must comply with applicable privacy and marketing legislation; they may not be distributed, transmitted, reproduced, published, licensed, transferred or sold in order to offer a similar or competing product; and an individual dataset may carry additional terms of its own. Insights produced by Bright Data may not be shared with third parties without prior written approval, and changes to a dataset need Bright Data's approval and may cost extra. On its own side, Bright Data uses account data to run and support the service, to communicate and market, and for statistical and analytical research; it processes publicly posted personal data on a legitimate interest basis in order to serve customers; and its CCPA notice states that identifiers may have been sold to third parties for commercial purposes over the previous twelve months.
Data retention & training
Hosting summary
Bright Data hosts on Amazon Web Services across multiple Availability Zones, with a disaster recovery site, encrypted backups taken every few minutes and distributed across several locations, and a recovery plan tested annually. Data in transit is protected with TLS 1.3, with support for at least TLS 1.2; data at rest uses AES-256 or stronger; credentials are salted and hashed. Independent auditors run annual application and infrastructure penetration tests, and a private bug bounty programme is in place. What the documentation never does is name a country or a region: no hosting jurisdiction, no EU or US region and no data centre location appears anywhere in the Trust Center. The company itself is established in Israel, and its privacy policy limits transfers outside the European Economic Area to countries the European Commission recognises as providing an adequate level of protection, or to recipients covered by contractual safeguards. Organisations with a data residency requirement will therefore need to obtain the hosting location in writing rather than from the published pages.
Things to keep in mind
Risks and trade-offs to weigh before adopting Bright Data.
- The Master Service Agreement lets Bright Data retain and reuse the data you collect at its sole discretion, with no documented opt-out: assume anything you scrape may also serve the vendor
- The CCPA notice states that identifiers may have been sold to third parties in the previous twelve months, which is worth reading closely before submitting personal data
- No hosting country, no region and no subprocessor list are published, so a data residency requirement cannot be satisfied from public documents alone
- No retention period is ever quantified; the policy relies on a purpose test rather than a clock, which makes deletion timelines impossible to plan
- Legal responsibility for the collection sits with you: the agreement makes the customer solely responsible for lawful grounds, notices and data subject rights when personal data is involved
- Bright Data may suspend access immediately at its sole discretion, subscription fees are non-cancellable and non-refundable, and liability is capped at one month of fees
- The agreement is governed by Israeli law with exclusive jurisdiction in the Tel-Aviv-Jaffa courts, class actions are waived, and the terms may be modified unilaterally and take effect on posting
Setup & Integrations
Technical difficulty
Two very different paths. Without code, ready-made datasets, no-code scrapers and the Managed Data Acquisition service need no engineering at all: the Web Scraper API is advertised as requiring no coding and no proxy management, and Scraper Studio launches scrapers in minutes with proxies, browsers and unblocking built in. With code, expect a conventional integration: REST API, Python and JavaScript SDKs, a CLI, and Selenium, Puppeteer or Playwright pointed at the Browser API. The real friction is administrative rather than technical, since several services only open after a Know Your Customer review.
Deployment
Integrations
Behind Bright Data
Fundraising
Social
Resources
All the official URLs gathered for verification and reference.
Alternatives
Tools that compete with or complement Bright Data.
Frequently asked questions
Is there a free plan, or only a trial?
What does Bright Data cost once you go paid?
Is there an API, and what languages are supported?
Can an AI agent use Bright Data directly?
Is Bright Data GDPR compliant?
Where is customer data hosted?
Who is behind Bright Data, and is there an age limit?
Is there a mobile app?
Should you pick Bright Data?
Bright Data is not a consumer AI tool but the infrastructure layer other data and AI products are built on, and its strength is coverage. One account opens a residential proxy network of more than 400 million IPs across 195 countries, scraper APIs for over 800 sites, a marketplace of pre-collected datasets, e-commerce intelligence and a Web MCP server that plugs an agent straight into the live web. Success-based billing and permanent free tiers make the platform genuinely testable before any commitment, and the compliance apparatus - ISO 27001, 27017, 27018 and 27701, SOC 2 Type II, SOC 3, CSA STAR, a PwC assurance report, Article 27 representatives in Brussels and London, and a published Data Protection Addendum - is more substantial than most vendors in this category put in public. The reservations are just as concrete. Pricing is fragmented across one page per product with incompatible units, so total cost is hard to model before a real workload runs, and the first committed monthly tier sits at USD 499. On the data side, the Master Service Agreement lets Bright Data retain and reuse whatever the customer collects at its sole discretion, no opt-out is documented, no subprocessor list is published, no hosting country is named beyond AWS with multiple Availability Zones, and no retention period is quantified; the CCPA notice acknowledges that identifiers may have been sold in the previous twelve months. The verdict follows from that split. For an engineering or AI team that needs public web data at industrial scale and can absorb a usage-based budget, this is one of the most complete options available. For anyone with strict data residency requirements, a need for a named subprocessor list, or a small and occasional workload, those gaps are worth settling in contract before committing.
- Choosing a selection results in a full page refresh.
- Opens in a new window.