---
title: "Bright Data MCP Server: review, setup & limitations · Undominated.ai"
canonical: https://undominated.ai/mcp-servers/brightdata/
description: "Search-engine results, page scraping, pre-built structured-data scrapers and a remote browser through Bright Data's proxy infrastructure, from a hosted URL or a local npx package."
---

# Bright Data MCP Server: review, setup & limitations · Undominated.ai

> Search-engine results, page scraping, pre-built structured-data scrapers and a remote browser through Bright Data's proxy infrastructure, from a hosted URL or a local npx package.

[← Explore all mcp servers](/mcp-servers/)

WEB EXTRACTION AND AUTOMATION / Bright Data

# Bright Data MCP Server

Search-engine results, page scraping, pre-built structured-data scrapers and a remote browser through Bright Data's proxy infrastructure, from a hosted URL or a local npx package.

 [See setup guidance ↓](#setup)[Original source ↗](https://github.com/brightdata/brightdata-mcp)

SOURCE REVIEW

 Reviewed 2026-10-13
 Evidence 3 linked sources
 Publisher Bright Data
 Licence [MIT ↗](https://github.com/brightdata/brightdata-mcp/blob/39cc2dcd9937512e22c369b81bad760333e2be61/LICENSE)
 Revision 39cc2dcd9937
 [Read what was—and wasn’t—checked ↓](#review)

“The Bright Data MCP server gives AI agents real-time access to public web data.”

 [Bright Data · upstream description ↗](https://github.com/brightdata/brightdata-mcp/blob/39cc2dcd9937512e22c369b81bad760333e2be61/README.md) Our analysis follows below.

01 / THE REASONING

## Why this made the selection.

 - Published in Bright Data's own repository; package.json at this commit names @brightdata/mcp version 3.0.0 with license MIT, matching the LICENSE (MIT, Copyright (c) 2025 Bright Data).
- The README documents a content gate, on by default, that asks the user before returning scraped content that reads like instructions to change the MCP configuration.

### A good fit for

 - Fetching pages and search results that block ordinary HTTP clients, and collecting ChatGPT and Perplexity answers as structured data
- Driving a remote browser session to navigate, click, type and take screenshots of pages

### Weigh up before choosing

 - The hosted URL carries the API token as a query parameter (?token=), where client logs can record it; the local npx form keeps it in API_TOKEN.
- The README says the content gate does not cover browser tools, and clients without MCP elicitation get matching content withheld.
- The README states a monthly free request allowance and pay-as-you-go rates; they were not checked against Bright Data's pricing page, and no request was made.

02 / THE REVIEW RECORD

## What we actually inspected.

Source review has boundaries. A clear record is more useful than a “safe” badge.

### Material inspected

 - README.md at 39cc2dcd9937512e22c369b81bad760333e2be61
- LICENSE at the same commit
- package.json at the same commit

### Our findings

 - The GROUPS and TOOLS environment variables, or groups= and tools= on the hosted URL, restrict which of the README's 32 tools are exposed.
- In the README's environment-variable table RATE_LIMIT defaults to unlimited and BASE_MAX_RETRIES to 0.
- Every request is routed through Bright Data's unblocking infrastructure, which the README says handles bot detection, CAPTCHAs and proxy rotation.

### Not established by this review

 - npx launch
- The hosted endpoint
- Any scrape, search or browser session
- The content gate

The review applies to the material and revision named here. A newer upstream release can change its behavior.

03 / PUT IT TO WORK

## Connect a server deliberately.

[Upstream setup instructions ↗](https://github.com/brightdata/brightdata-mcp/blob/39cc2dcd9937512e22c369b81bad760333e2be61/README.md)

DOCUMENTED COMMAND

 npx @brightdata/mcp Copy command ↗

Copying does not execute this command. It may retrieve a newer version than the reviewed source.

 - The README names the line above as the local instance; register it in the client with API_TOKEN in env. It was not run.
- Set GROUPS or TOOLS to expose only the tools you need. Do not commit the token.

### Before you start

 - Node.js with npx
- A Bright Data API token

### Documented configuration example

Use the client and authentication context described in the setup guidance. Configuration formats vary between hosts.

 {
 "mcpServers": {
 "Bright Data": {
 "command": "npx",
 "args": [
 "@brightdata/mcp"
 ],
 "env": {
 "API_TOKEN": "your-token-here"
 }
 }
 }
} Copy configuration ↗

### Compatibility

Claude Desktop · Claude Code · Cursor · VS Code · Windsurf · Gemini CLI · Zed · Warp

### Implementation

JavaScript

### Transports

Streamable HTTP · stdio

### Hosted endpoint or local process with a Bright Data API token

 - Sends URLs, search queries and browser actions through Bright Data's proxy and unblocking infrastructure
- Runs pre-built scrapers and remote browser sessions billed to the account

### Cost model

MIT server source. The README states a recurring monthly free tier and pay-as-you-go pricing; no rate was verified.

04 / FOLLOW THE EVIDENCE

## The source trail.

Our notes are separate from the original resource. Check upstream before adopting a new version.

 - [Bright Data MCP README ↗](https://github.com/brightdata/brightdata-mcp/blob/39cc2dcd9937512e22c369b81bad760333e2be61/README.md) Checked 2026-10-13 https://github.com/brightdata/brightdata-mcp/blob/39cc2dcd9937512e22c369b81bad760333e2be61/README.md Supports: Hosted and local setup, Content gate, Environment variables, Pricing statement
- [Bright Data MCP LICENSE ↗](https://github.com/brightdata/brightdata-mcp/blob/39cc2dcd9937512e22c369b81bad760333e2be61/LICENSE) Checked 2026-10-13 https://github.com/brightdata/brightdata-mcp/blob/39cc2dcd9937512e22c369b81bad760333e2be61/LICENSE Supports: MIT License, Copyright (c) 2025 Bright Data
- [Bright Data MCP package.json ↗](https://github.com/brightdata/brightdata-mcp/blob/39cc2dcd9937512e22c369b81bad760333e2be61/package.json) Checked 2026-10-13 https://github.com/brightdata/brightdata-mcp/blob/39cc2dcd9937512e22c369b81bad760333e2be61/package.json Supports: Package @brightdata/mcp 3.0.0, License MIT

KEEP COMPARING

## Other approaches to consider.

Related by category or shared topics. These are alternatives to inspect, not a measured quality order.

 [### Apify MCP Server ↗ Searches Apify Store and runs Apify Actors (scrapers, crawlers and automation tools) from an MCP client, through Apify's hosted endpoint or a local stdio package.](/mcp-servers/apify/)[### Firecrawl MCP ↗ Provides web scraping, search, crawling and browser-interaction jobs through Firecrawl’s API.](/mcp-servers/firecrawl/)[### Algolia MCP Server ↗ Algolia's managed MCP offerings: Public MCP exposes application-scoped search and recommendations on selected indices for customer-facing agents; Productivity MCP gives user-scoped, read-only exploration and analytics; DevEx MCP serves version-matched API client snippets to coding agents without touching data.](/mcp-servers/algolia/)

 [AI Tools ↗](/tools/)[Skills ↗](/skills/)[Agents ↗](/agents/)[MCP Servers ↗](/mcp-servers/)[Workflows ↗](/workflows/)

## Continue your investigation

 - [Inspect resource evidence](/tools/)
- [Choose a host](/agents/)
- [Inspect reusable workflows](/skills/)
- [Use Undominated data](/api/)
