How to Set Up SE Ranking MCP for AI-Powered SEO (141 Tools)
SE Ranking's MCP server gives your AI agent access to 141 SEO tools — keyword research, domain analysis, backlink audits, SERP tracking, website audits, and something most SEO tools don't have yet: AI search visibility tracking. You can ask your agent "How visible is my brand in ChatGPT responses?" and get real data back.
We set this up on our own site (teamday.ai) alongside Ahrefs MCP and tested it end-to-end. Here's the complete guide — including what SE Ranking does better than Ahrefs, where Ahrefs still wins, and how to use both together.
What SE Ranking MCP Gives Your AI Agent
The server exposes two API categories, each with a dedicated token. You can use one or both.
Data API (70+ tools)
Research and analysis tools. No projects needed.
- ● Keyword Research — similar, related, long-tail, question keywords with volume, CPC, difficulty
- ● Domain Analysis — organic/paid keywords, competitors, traffic history, page rankings
- ● Backlinks — full backlink index, referring domains, anchors, authority scores
- ● SERP Analysis — organic results, ads, featured snippets, AI Overview
- ● Website Audit — crawl your site, find technical issues, track fixes
- ★ AI Search Visibility — track your brand in ChatGPT, Perplexity, and other AI engines
Project API (60+ tools)
Project management and rank tracking.
- ● Rank Tracking — monitor keyword positions daily, track competitors
- ● Competitor Monitoring — add competitors, track their rankings, compare positions
- ● Backlink Monitor — track specific backlinks, manage disavow lists
- ● Marketing Plan — manage SEO tasks and action items
- ● Project Management — create projects, manage keywords, configure search engines
- ● Search Console Integration — pull GSC data through SE Ranking's interface
The standout feature: AI Search Visibility tracking is something most SEO tools don't offer yet. SE Ranking tracks how often your brand appears in AI-generated responses from ChatGPT, Perplexity, and other LLM-powered search engines — and which prompts trigger those mentions. This is becoming critical as more users shift from Google to AI search.
SE Ranking vs Ahrefs MCP: When to Use Which
We run both SE Ranking and Ahrefs MCPs simultaneously. They're complementary, not redundant. Here's how they compare:
| Capability | SE Ranking | Ahrefs |
|---|---|---|
| Keyword research | Strong — similar, related, long-tail, questions | Strong — matching terms, related terms, suggestions |
| Backlink index | Good | Largest backlink index in the industry |
| Domain authority metric | Domain InLink Rank | Domain Rating (DR) — industry standard |
| Website audit | Full crawl via API — with JS rendering option | Site Audit via projects (requires Ahrefs project) |
| AI search visibility | Yes — ChatGPT, Perplexity, AI Overview tracking | Brand Radar — LLM mentions and share of voice |
| SERP analysis | Full SERP dump — organic, ads, AI Overview, maps, reviews | Top positions with backlink metrics per result |
| Keyword gap analysis | Built-in comparison tool between two domains | Manual via organic keywords + filtering |
| Rank tracking | Full project-based rank tracking via API | Rank Tracker with competitor comparison |
| Pricing | More affordable, especially for startups | Premium pricing, larger data index |
Our recommendation: Use Ahrefs when you need the deepest backlink data and domain rating benchmarks. Use SE Ranking for keyword gap analysis, website audits, SERP analysis (especially AI Overview data), and AI search visibility tracking. Running both gives your AI agent the most complete SEO picture.
What You Need Before Starting
- ✓ An SE Ranking account — sign up here (paid plan required for API access)
- ✓ An MCP-compatible AI client — TeamDay (easiest — no Docker required), Claude Code, Claude Desktop, or Gemini CLI
TeamDay users: You don't need Docker or any local setup. TeamDay runs the SE Ranking MCP server for you — just paste your API token and go. If you're using Claude Code or another client directly, you'll need Docker installed.
Setup Guide
Unlike Google Analytics MCP (which requires service accounts, API enablement, and cross-system permissions), SE Ranking's setup is refreshingly simple: get a token, connect it.
Get Your API Token
SE Ranking provides two types of API tokens. You can use one or both, depending on what you need:
Data API Token
UUID format: 80cfee7d-xxxx-xxxx-xxxx-fc8500816bb3
For keyword research, domain analysis, backlinks, SERP, audits, AI visibility.
Project API Token
Hex format: 253a73adxxxx...340aa0a939
For rank tracking, competitor monitoring, backlink management, marketing plans.
- Log in to your SE Ranking account
- Go to API Dashboard
- Click Create Token
- Give it a name (e.g., "TeamDay MCP")
- Copy the token — you'll need it in Step 2
Note: Data API tokens expire after 14 days by default. You can create new ones anytime from the API Dashboard. Keep your token secure — it grants read access to SE Ranking data.
Connect to Your AI Agent
Option A: TeamDay (Recommended)
TeamDay has SE Ranking built into the MCP library. No Docker, no config files, no terminal.
- Open your Space in TeamDay
- Click MCP Servers in the sidebar
- Find SE Ranking in the Library tab (under Marketing)
- Paste your API token and click Connect
That's it. TeamDay encrypts your token, manages the server lifecycle, and your AI agent gets all 141 SE Ranking tools immediately. No restarts needed.
Option B: Claude Code / Claude Desktop / Gemini CLI
For direct CLI/desktop use, SE Ranking's MCP runs as a Docker container. First build the image:
Then add to your project's .mcp.json
or Claude Desktop's claude_desktop_config.json:
{
"mcpServers": {
"se-ranking": {
"command": "docker",
"args": [
"run", "-i", "--rm",
"-e", "DATA_API_TOKEN",
"-e", "PROJECT_API_TOKEN",
"se-ranking/seo-data-api-mcp-server"
],
"env": {
"DATA_API_TOKEN": "your-data-api-token-here",
"PROJECT_API_TOKEN": "your-project-api-token-here"
}
}
}
}
If you only have a Data API token, omit the PROJECT_API_TOKEN
lines. The server handles missing tokens gracefully — PROJECT_ tools return an error while DATA_ tools work normally.
Config file locations:
- Claude Code (project):
.mcp.jsonin your project root - Claude Desktop (macOS):
~/Library/Application Support/Claude/claude_desktop_config.json - Claude Desktop (Windows):
%AppData%\Claude\claude_desktop_config.json - Gemini CLI:
~/.gemini/settings.json
Tip: Store the cloned repo somewhere persistent
(e.g., ~/mcps/se-ranking-mcp/)
so you can easily update later with git pull && docker build.
Verify It Works
After adding the configuration, restart your MCP client. Then test with a simple query:
If you see real data come back, you're all set. If you get an authentication error, double-check that your token is correctly pasted in the config (no extra spaces or quotes).
Quick test with Docker directly:
echo '{"jsonrpc":"2.0","method":"initialize","params":{"protocolVersion":"2024-11-05","capabilities":{},"clientInfo":{"name":"test","version":"1.0"}},"id":1}' \
| docker run -i --rm \
-e DATA_API_TOKEN=your-token-here \
se-ranking/seo-data-api-mcp-server
You should see a JSON response with protocolVersion
and serverInfo. That confirms the MCP server
is running and accepting connections.
What You Can Ask Your AI Agent
Here are real examples of what becomes possible once SE Ranking MCP is connected:
Keyword Research
"Find me 20 long-tail keywords related to 'AI agent platform' with search volume over 100 and keyword difficulty under 40."
Competitor Analysis
"Compare teamday.ai with relevanceai.com — what keywords do they rank for that we don't? Sort by volume, show the top 30 gaps."
SERP Analysis
"Analyze the SERP for 'claude code tutorial'. Show me the top 10 organic results, any featured snippets, and whether there's an AI Overview."
AI Search Visibility
"How visible is 'TeamDay' in AI search engines? What prompts trigger mentions of our brand in ChatGPT and Perplexity?"
Website Audit
"Run a technical SEO audit on teamday.ai. Find broken links, missing meta tags, slow pages, and any crawl issues."
Backlink Analysis
"Show me the top 20 referring domains for teamday.ai. What's our domain authority? Which pages get the most backlinks?"
Built-in Prompt Templates
SE Ranking's MCP server includes 5 pre-built prompt templates for complex multi-step analysis:
- serp-analysis — Compare SERPs across two locations for the same keyword
- backlink-gap — Find backlink opportunities from competitors you're missing
- domain-traffic-competitors — Full domain traffic and competitor landscape report
- keyword-clusters — Generate keyword clusters with intent grouping and content ideas
- ai-share-of-voice — Compare brand visibility across AI search engines vs competitors
Combining SE Ranking with Other MCP Servers
The real power emerges when your AI agent can cross-reference data from multiple SEO sources in a single conversation:
SE Ranking + Ahrefs
Use SE Ranking for keyword gaps and AI visibility, Ahrefs for backlink depth and Domain Rating. Two data sources, complete picture.
SE Ranking + Search Console
Cross-reference SE Ranking's keyword data with your actual Google impressions and clicks. Find keywords where you rank but aren't getting clicks.
SE Ranking + Google Analytics
Map keyword rankings to actual traffic and conversions. Know which SEO improvements drive real business results.
Troubleshooting
"Docker: no space left on device"
Cause: Docker's virtual disk is full from old images and containers.
Fix: Run docker system prune -f && docker builder prune -f
to free space. We freed 42GB doing this before our build succeeded.
"Authentication error" from DATA_ tools
Cause: Token is expired, invalid, or has extra whitespace.
Fix: Data API tokens expire after 14 days. Generate a new one at SE Ranking API Dashboard. Make sure there are no trailing spaces in your config.
PROJECT_ tools return errors but DATA_ tools work
Cause: You only configured a Data API token, not a Project API token.
Fix: This is expected if you only use Data API features. To use
Project tools (rank tracking, competitor monitoring), add a
PROJECT_API_TOKEN
to your config.
MCP server not appearing after restart
Cause: Docker daemon isn't running, or the image name is wrong.
Fix: Check that Docker Desktop is running. Verify the image exists with
docker images | grep se-ranking.
The image name must exactly match se-ranking/seo-data-api-mcp-server.
Rate Limits and API Usage
Data API
10 requests per second (default)
Project API
5 requests per second (default)
These limits are generous for interactive AI conversations. If you're running automated batch workflows and need higher limits, contact [email protected].
Ready to Supercharge Your SEO?
TeamDay connects SE Ranking, Ahrefs, Search Console, and Google Analytics into one AI-powered SEO workspace. Your agent does the analysis — you make the decisions.
Try TeamDay FreeNo credit card required. Connect your SEO tools in minutes.