DataForSEO MCP Server
Official DataForSEO MCP server: query SERP data, keyword volumes, backlinks and on-page analyses right inside your AI assistant, hosted with OAuth or locally via npx.
- Skill Road
- DataForSEO MCP Server
Description
The DataForSEO MCP Server connects your AI assistant to the complete DataForSEO data interface. You can query search engine results pages (SERPs), keyword search volumes, backlink profiles, on-page checks and competitor data in plain language without writing a single line of API code yourself. MCP, the Model Context Protocol, is an open standard that assistants such as Claude Code, Claude Desktop or Cursor use to call external tools. The server comes from DataForSEO itself, is published on GitHub as an open source project under the Apache 2.0 licence and ships as the npm package dataforseo-mcp-server.
What is new in version 3?
According to the README, the current release is a fundamentally new version. The earlier server, with many separate tools per API area, is deprecated and has moved to a repository of its own. Version 3 gets by with just four tools: three of them search the DataForSEO API documentation, the fourth sends authenticated requests to any endpoint. So the assistant first looks up which endpoint fits your question and which parameters it needs, then calls it in a targeted way. That keeps the tool list in your client lean while still covering the entire API. According to the vendor, the documentation is cached for 24 hours.
What can you do with it?
Typical questions include: Which ten pages rank in Germany for a given keyword? What is the monthly search volume for a list of keywords? Which referring domains has a competitor recently gained? Which technical issues does an on-page check find on a specific URL? The assistant fetches the data, summarises it and can turn it into tables, content ideas or priority lists. For SEO agencies and in-house teams that used to copy data together from several tools, this saves a lot of manual work.
Hosted or local
DataForSEO runs a public remote server at https://mcp.dataforseo.com/v3/mcp. Clients with HTTP support discover the OAuth sign-in on their own, so you never type your credentials into a configuration file. Alternatively, you start the server locally with npx. In local stdio mode OAuth is not used; instead you provide your API login and API password as environment variables. A local HTTP mode on port 3000 is available as well.
Trim responses and keep an eye on cost
Many DataForSEO responses are very large. According to the README, a field configuration lets you decide which fields each endpoint returns. That saves context in the AI model and makes answers easier to read. Important for cost: every request through the MCP server is an ordinary API request and is billed exactly like direct usage. So do not let the assistant fire off bulk queries unchecked.
Cost
The server itself is free and open source. For the data you need a DataForSEO account. According to the pricing page, DataForSEO works on a pay-as-you-go basis without a subscription, new accounts receive a small starting credit for testing, and for regular use you top up your balance. Current terms are listed on the official pricing page.
Who is the DataForSEO MCP Server for?
It pays off for anyone who wants flexible access to SEO data without a fixed monthly plan: freelancers, agencies with fluctuating data needs and developers who build their own SEO tools. If you already pay for an all-in-one suite such as Ahrefs or Semrush, you get a finished interface there; DataForSEO stands out instead with the breadth of its raw data and usage-based billing. Start with the trial credit and simple queries before you kick off larger analyses.
Requirements
A DataForSEO account with API access (API login and API password at app.dataforseo.com/api-access). For the hosted server, an MCP client with HTTP transport and OAuth, such as Claude Code or Cursor. For local use, additionally Node.js version 20 or later.
Installation instructions
Step 1: Create a DataForSEO account
Sign up at dataforseo.com. You will then find your API credentials at https://app.dataforseo.com/api-access. Important: the API login and API password are not the same as your regular dashboard password. You do not need these credentials for the hosted server with OAuth, but you do for local use.
Step 2: Add it to Claude Code (recommended)
Open a terminal and run this command. It registers the hosted DataForSEO server:
claude mcp add --transport http dataforseo https://mcp.dataforseo.com/v3/mcp
Then start Claude Code and enter the following command. Select dataforseo and click Authenticate. In the browser, sign in to DataForSEO and approve access.
/mcp
Step 3: Alternatively add it to Cursor or Claude Desktop
In Cursor, open the file ~/.cursor/mcp.json (in Claude Desktop, go to Settings, Developer, Edit Config to open claude_desktop_config.json) and add this entry:
{
"mcpServers": {
"dataforseo": {
"url": "https://mcp.dataforseo.com/v3/mcp"
}
}
}
Save the file and fully restart the client. On first use, the OAuth sign-in opens in your browser.
Step 4: Alternatively run it locally with your API login
If your client does not support OAuth or you want to run the server locally, first check that Node.js version 20 or later is installed:
node --version
Then add the server with your credentials to your client's configuration file. Replace YOUR_API_LOGIN and YOUR_API_PASSWORD with the values from step 1:
{
"mcpServers": {
"dataforseo": {
"command": "npx",
"args": ["-y", "dataforseo-mcp-server"],
"env": {
"DATAFORSEO_LOGIN": "YOUR_API_LOGIN",
"DATAFORSEO_PASSWORD": "YOUR_API_PASSWORD"
}
}
}
}
In Claude Code you can also do this with a single command:
claude mcp add dataforseo --env DATAFORSEO_LOGIN=YOUR_API_LOGIN --env DATAFORSEO_PASSWORD=YOUR_API_PASSWORD -- npx -y dataforseo-mcp-server
Step 5: Test the connection
Ask your assistant a simple question, for example:
Use DataForSEO and show me the top 10 Google results in the United States for the keyword "solar panels".
If the results appear, everything works. Afterwards, check in the DataForSEO dashboard how much credit the request used so you get a feel for the cost.
Step 6: Trim responses (optional)
For local use, a field configuration lets you decide which fields an endpoint returns. Create a file named field-config.json:
{
"supported_fields": {
"/v3/serp/google/organic/live/advanced": ["items.title", "items.url", "items.rank_absolute"]
}
}
In the configuration from step 4, append the value "--configuration" followed by the absolute path to this file to the args.
claude mcp add --transport http dataforseo https://mcp.dataforseo.com/v3/mcp
Authentication
Hosted server and local HTTP mode: OAuth 2.0, the client handles the sign-in in the browser. Local stdio mode: API login and API password in DATAFORSEO_LOGIN and DATAFORSEO_PASSWORD (HTTP Basic).
Required access permissions
Reading the API documentation and authenticated requests to every DataForSEO endpoint your account has access to, including SERP, keywords, backlinks, on-page, domain analytics and content analysis. Every request uses credit.
Transmitted or stored data
Keywords, domains and URLs from your requests are sent to the DataForSEO API, and with the hosted server additionally through mcp.dataforseo.com. The returned data enters the AI client context and therefore reaches the model provider.
Security risks
An assistant can trigger many paid requests without asking and burn through credit. In local mode, the API login and API password sit in plain text in the client configuration.
License and costs
- License
- Apache-2.0
- Cost
- paid
The server is open source (Apache 2.0). According to the pricing page, data is billed by usage without a subscription; new accounts get a small starting credit, after which you top up your balance.
Alternatives
Not recorded yet.
At a glance
- Provider
- DataForSEO
- Status
- Official server
- Deployment
- Local and remote
- Current version
- Not recorded yet.
- Last reviewed
- 27.10.2026
Repository and documentation
Categories
Supported clients
Related guides
Guides and background related to this entry.
Set up and use Cookbook Audit safely
How Anthropic's official Cookbook Audit skill automatically checks and reviews Jupyter notebooks against a fixed scoring rubric.
09.09.2026
Use Source Management safely
Guide to transparent enterprise search: identify sources, set clear priorities, explain rate limits, and limit risky data access.
09.09.2026
Research Microsoft Docs safely and traceably
GitHub's skill routes queries about Azure, .NET, VS Code, and GitHub to the correct official Microsoft documentation source.
09.09.2026
Set up Content Creation safely
The Content Creation skill provides templates and best practices for blog posts, social media, email, landing pages, and case studies.
09.09.2026