# Bright Data

> Search, scrape, and unblock any public web page from an AI agent — official Web Unlocker MCP.

[Canonical HTML page](https://top-mcps.com/mcp/bright-data) · [server.json](https://top-mcps.com/mcp/bright-data.json) · [methodology](https://top-mcps.com/about/methodology)

## Install

### Claude Desktop — `claude_desktop_config.json`

Paste under mcpServers. Fully quit and reopen Claude after editing.

```json
{
  "mcpServers": {
    "bright-data": {
      "command": "npx",
      "args": [
        "-y",
        "@brightdata/mcp"
      ],
      "env": {
        "API_TOKEN": "${API_TOKEN}"
      }
    }
  }
}
```

### Claude Code — `CLI or .mcp.json`

Run from your repo. Commit .mcp.json to share with your team.

```shell
# export API_TOKEN=your_brightdata_token
claude mcp add bright-data -- npx -y @brightdata/mcp
```

### Cursor — `.cursor/mcp.json`

Global path: ~/.cursor/mcp.json. Reload window after editing.

```json
{
  "mcpServers": {
    "bright-data": {
      "command": "npx",
      "args": [
        "-y",
        "@brightdata/mcp"
      ],
      "env": {
        "API_TOKEN": "${API_TOKEN}"
      }
    }
  }
}
```

### VS Code — `.vscode/mcp.json`

VS Code uses the "servers" key (not "mcpServers").

```jsonc
{
  "servers": {
    "bright-data": {
      "command": "npx",
      "args": [
        "-y",
        "@brightdata/mcp"
      ],
      "env": {
        "API_TOKEN": "${API_TOKEN}"
      }
    }
  }
}
```

### Windsurf — `~/.codeium/windsurf/mcp_config.json`

Open via Cascade → hammer icon → Configure.

```json
{
  "mcpServers": {
    "bright-data": {
      "command": "npx",
      "args": [
        "-y",
        "@brightdata/mcp"
      ],
      "env": {
        "API_TOKEN": "${API_TOKEN}"
      }
    }
  }
}
```

### Cline — `cline_mcp_settings.json`

Open via the Cline sidebar → MCP Servers → Edit.

```json
{
  "mcpServers": {
    "bright-data": {
      "command": "npx",
      "args": [
        "-y",
        "@brightdata/mcp"
      ],
      "env": {
        "API_TOKEN": "${API_TOKEN}"
      }
    }
  }
}
```

### Continue — `~/.continue/config.json`

Continue uses modelContextProtocolServers with a transport block.

```json
{
  "experimental": {
    "modelContextProtocolServers": [
      {
        "transport": {
          "type": "stdio",
          "command": "npx",
          "args": [
            "-y",
            "@brightdata/mcp"
          ],
          "env": {
            "API_TOKEN": "${API_TOKEN}"
          }
        }
      }
    ]
  }
}
```

### Codex CLI — `~/.codex/config.toml`

Codex uses TOML. Each server is a [mcp_servers.<name>] subtable.

```shell
# ~/.codex/config.toml
[mcp_servers.bright-data]
command = "npx"
args = [
  "-y",
  "@brightdata/mcp",
]
env = { API_TOKEN = "${API_TOKEN}" }
```

### Zed — `~/.config/zed/settings.json`

Zed calls them "context_servers". Settings live-reload on save.

```jsonc
{
  "context_servers": {
    "bright-data": {
      "command": {
        "path": "npx",
        "args": [
          "-y",
          "@brightdata/mcp"
        ]
      },
      "env": {
        "API_TOKEN": "${API_TOKEN}"
      }
    }
  }
}
```

### ChatGPT — `ChatGPT → Apps directory`

Bright Data doesn't ship a hosted HTTPS endpoint today. ChatGPT supports remote MCP servers only — to use this server in ChatGPT you'll need to deploy it to a public HTTPS URL first (e.g. via Cloudflare Workers or Vercel) or wait for an official remote build.

```none

```

## At a glance

- **Maintainer:** Bright Data
- **Transport:** stdio
- **Auth model:** API key
- **Required secrets:** API_TOKEN
- **Supported clients:** Claude, Cursor, VS Code, Any MCP-compatible client, Bright Data accounts
- **License:** MIT
- **Language:** JavaScript
- **Latest version:** latest
- **Last verified:** 2026-07-16
- **Score:** 67/100 (rubric 2026-04 — see https://top-mcps.com/about/methodology)
- **Source:** https://github.com/brightdata/brightdata-mcp

## Quick answer

**What it does.** Exposes Bright Data operations as MCP tools: web search, single-page scrape via the Web Unlocker (anti-bot, CAPTCHA solving, geo-routing), and discovery search over the open web.

**Best for:**
- Scraping sites with aggressive bot detection
- Geo-targeted page fetches
- Search + unblockable fetch in one MCP
- Bypassing CAPTCHAs and rate blocks
- High-volume public web access

**Not for:**
- Sites with an existing API
- Low-volume fetches Fetch/Firecrawl handle
- Workflows that cannot use a paid data network

## Description

Bright Data's official MCP (MIT, 2.5k+ stars) gives an agent unblockable public web access: search engines, single-page scrape through the Web Unlocker (which handles CAPTCHAs, bot detection, and geo-blocks), and AI-ranked discovery. It is the pick when a target actively fights scrapers and Firecrawl or a raw fetch gets blocked.

## Why it matters

Most scraping MCPs fail on hard targets — sites with aggressive bot detection or geo-walls. Bright Data's network is built to get through those, so an agent gets the page instead of a 403 or a CAPTCHA.

## Key features

- Web Unlocker: CAPTCHA + bot-detection bypass
- Search + scrape + discovery tools
- 5,000 free credits/month
- stdio via npx, API-token auth
- Official Bright Data, MIT

## FAQ

### How is it different from Firecrawl?

Firecrawl is excellent for clean markdown on cooperative sites; Bright Data's Web Unlocker is built for hostile targets — CAPTCHAs, bot detection, and geo-blocks. Reach for Bright Data when Firecrawl gets a 403.

### What does it cost?

Accounts get 5,000 credits/month free (no card required), then usage-based pricing. The MCP itself is MIT-licensed and free.

### Is there a Pro mode?

Yes — Pro mode unlocks additional structured-data and browser tools beyond the default search + scrape surface.

## Changelog

- **2026-07-16** — Refreshed install snippets and fact sheet; verified for 2026.
- **2025-04-15** — Initial directory listing.