Get instant domain-level categorization lookups from a pre-classified database of 30 million or more domains. Connect your AI assistant directly to our database via the Model Context Protocol and retrieve IAB categories, web filtering data, buyer personas, and PageRank scores in milliseconds.
The Model Context Protocol is an open standard developed by Anthropic that enables AI assistants to communicate with external tools, databases, and APIs in a structured, secure way.
The Model Context Protocol, commonly known as MCP, was created to solve a fundamental challenge in the AI ecosystem: how do you give large language models reliable, real-time access to external data and functionality without compromising security or requiring complex custom integrations? Before MCP, every tool integration required bespoke code, API wrappers, and fragile prompt engineering. MCP changes this by providing a universal communication layer between AI applications and the tools they need.
At its core, MCP follows a client-server architecture. The AI application, whether that is Claude Desktop, Claude Code, Cursor, or any other MCP-compatible assistant, acts as the client. The MCP server runs locally on your machine and exposes a set of tools that the AI can discover and call. Communication happens over standard input and output streams using a JSON-based protocol, which means the server process stays contained within your local environment and never needs to be exposed to the internet.
When you ask your AI assistant a question like "What category is stripe.com?", the assistant recognizes that it needs external data. It sends a structured tool call to the MCP server. The server processes that request, connects to the websitecategorizationapi.com Database API, retrieves the categorization data, and returns the result to the assistant. The entire round trip happens in milliseconds for database lookups, making the experience feel seamless and instantaneous.
Because MCP servers run locally, your API key stays on your machine and is never transmitted to the AI model provider. The server only communicates with the specific API endpoints it was designed to use. This architecture gives you full control over what tools are available, how they are configured, and what data flows through them. You can inspect every request and response, stop the server at any time, and configure it to fit your specific workflow.
The Database Lookup MCP server connects your AI assistant directly to our pre-classified database of over 30 million domains, providing instant categorization results for domain-level queries.
This MCP server is designed for speed and efficiency. Instead of analyzing webpages in real time, it retrieves pre-computed categorization data from our database, which means lookups return in milliseconds rather than seconds. Every domain in the database has already been crawled, analyzed, and classified across multiple taxonomy standards, so the data you receive is comprehensive and immediately usable.
The Database Lookup server supports domain-level queries only. You can look up domains like
stripe.com or bbc.co.uk, but you cannot look up individual subpages such as
bbc.co.uk/news/technology. If you need subpage-level classification, use the
Real-Time API MCP server instead.
Each domain record in the database contains a rich set of categorization signals and metadata. When you look up a domain, the server returns all of the following data fields, giving you a multi-dimensional understanding of any website in a single request:
Full IAB Content Taxonomy version 2 classifications across all four tiers of depth. This is the industry standard for digital advertising and content categorization, ranging from broad top-level categories down to granular sub-categories that precisely describe the domain's content focus.
The latest IAB Content Taxonomy version 3 classifications with updated category structure and expanded coverage. Version 3 introduces new categories for emerging content types and refines the existing hierarchy, ensuring your data stays aligned with current industry standards.
A content-filtering classification designed for security, compliance, and parental control applications. This identifies domains by content type in categories such as news, social media, adult content, gambling, malware, and dozens more, enabling automated policy enforcement.
Audience and buyer persona classifications that describe the typical visitors and customers of each domain. This data is useful for ad targeting, lead enrichment, and market research, helping you understand who is likely interacting with a given website.
A domain authority metric based on the Open PageRank initiative. This numerical score indicates the relative importance and authority of each domain on the web, useful for filtering out low-quality sites and prioritizing high-traffic, reputable domains in your workflows.
The primary country associated with each domain, determined through a combination of hosting location, content language, top-level domain, and audience geo-distribution analysis. This enables geographic filtering, regional compliance checks, and market segmentation.
The database contains over 30 million pre-classified domains, covering an estimated 99% or more of active internet traffic by volume. The most popular domains, long-tail niche sites, and regional web properties are all represented, making it suitable for the vast majority of domain-level categorization needs.
The Database Lookup MCP server exposes five tools that your AI assistant can discover and call automatically. Each tool is designed for a specific type of query.
lookup_domain Single Domain LookupThis is the most commonly used tool in the server. It takes a single domain name as input and returns the complete categorization record from the database. The response includes IAB v2 and v3 categories across all tiers, the web filtering classification, buyer persona data, the OpenPageRank score, and country information. Use this when you want detailed intelligence on one specific domain.
Example conversation with your AI assistant:
// You ask Claude:
"Look up the domain stripe.com in the database."
// Claude calls lookup_domain with domain="stripe.com"
// and returns something like:
Domain: stripe.com
IAB v2 Tier 1: Technology & Computing
IAB v2 Tier 2: Computing > Internet Technology
Web Filtering: Financial Services
Buyer Persona: Tech-Savvy Business Decision Maker
PageRank: 8.42
Country: USbulk_lookup Batch Domain Lookup (up to 100)The bulk lookup tool accepts an array of up to 100 domain names in a single request and returns the full categorization record for each domain found in the database. This is significantly more efficient than calling lookup_domain repeatedly, both in terms of network overhead and credit consumption time. Any domains not found in the database will return null values, which indicates you should fall back to the Real-Time API for those specific domains.
Example conversation:
// You ask Claude:
"Look up these domains in bulk: google.com, stripe.com,
shopify.com, github.com, and notion.so"
// Claude calls bulk_lookup with all 5 domains
// Returns categorization data for each domainsearch_domains Search by Category or CountryThis tool enables reverse lookups: instead of querying a known domain, you can search the database by category or country to discover domains that match specific criteria. For example, you can find all domains classified as "Financial Services" in Germany, or search for e-commerce domains in the United Kingdom. This is useful for market research, competitive analysis, lead generation, and building targeted domain lists.
// You ask Claude:
"Search the database for e-commerce domains based in Germany."
// Claude calls search_domains with the appropriate parameters
// Returns a list of matching domains with their categorization datadatabase_stats View Database StatisticsReturns high-level statistics about the database, including the total number of classified domains, category distribution breakdowns, geographic coverage information, and the timestamp of the most recent database update. This tool takes no input parameters and is useful for understanding the scope and freshness of the data you are working with.
// You ask Claude:
"Show me the database statistics."
// Claude calls database_stats
// Returns total domain count, category coverage, last update date, etc.check_credits Monitor Your UsageQueries your account to display the current credit balance and usage information. Each domain lookup (single or within a bulk request) costs one credit. This tool helps you monitor consumption and plan ahead, especially when running large batch operations. It takes no input parameters and returns your remaining credit balance alongside recent usage data.
// You ask Claude:
"How many credits do I have left?"
// Claude calls check_credits and returns your balanceFollow these steps to install and configure the Database Lookup MCP server on your machine. The entire setup process typically takes about five minutes.
If you do not already have an account, go to websitecategorizationapi.com and create a free account. Registration is quick and only requires an email address. Once your account is active, you will have access to the API dashboard and credit management system. New accounts include a set of free trial credits so you can test the service before purchasing additional credits.
After logging in, navigate to the Profile page from your dashboard. Your API key is displayed in the API credentials section. Copy this key and keep it in a secure location. You will need this key to authenticate the MCP server when it connects to the database API. Treat your API key like a password and do not share it publicly or commit it to version control repositories.
The MCP server is built in Python and requires version 3.10 or higher. Check your current Python version by opening a terminal and running the command below. If you need to install or upgrade Python, visit python.org/downloads for the latest installer. On macOS, you can also install Python via Homebrew with brew install python. On Ubuntu or Debian-based Linux distributions, use sudo apt install python3.
python3 --version
# Should output: Python 3.10.x or higher
Download the Database Lookup MCP server package from the provided link or clone the repository. The server consists of a small set of Python files and a requirements file. Place the files in a directory where you want the server to live. A common location would be a directory within your home folder, such as ~/mcp-servers/database-lookup/. The directory should contain the main server script, the requirements.txt file, and a template for the environment configuration.
# Create a directory and navigate into it
mkdir -p ~/mcp-servers/database-lookup
cd ~/mcp-servers/database-lookup
# Copy or download the server files here
The server depends on three Python packages: mcp (the Model Context Protocol SDK for Python), httpx (a modern async HTTP client), and python-dotenv (for loading environment variables from a .env file). Install all dependencies at once using the included requirements file. It is recommended to use a virtual environment to keep these packages isolated from your system Python installation.
# Optional: create and activate a virtual environment
python3 -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate
# Install the required packages
pip install -r requirements.txt
# The requirements.txt contains:
# mcp
# httpx
# python-dotenv
The server needs your API key to authenticate requests. The recommended approach is to create a .env file in the server directory with your key. Alternatively, you can set an environment variable in your shell profile. Either method works, but the .env file approach is simpler and keeps the configuration contained within the server directory.
# Create a .env file in the server directory
WEBSITE_CATEGORIZATION_API_KEY=your_api_key_here# Add to ~/.bashrc, ~/.zshrc, or equivalent
export WEBSITE_CATEGORIZATION_API_KEY="your_api_key_here"
Claude Desktop discovers MCP servers through its configuration file. The location of this file differs by operating system. Open the file in a text editor and add the server configuration in the mcpServers object. If the file does not exist, create it. After saving the configuration, restart Claude Desktop to load the new server.
{
"mcpServers": {
"website-categorization-db": {
"command": "python3",
"args": [
"/Users/YOUR_USERNAME/mcp-servers/database-lookup/server.py"
],
"env": {
"WEBSITE_CATEGORIZATION_API_KEY": "your_api_key_here"
}
}
}
}{
"mcpServers": {
"website-categorization-db": {
"command": "python",
"args": [
"C:\\Users\\YOUR_USERNAME\\mcp-servers\\database-lookup\\server.py"
],
"env": {
"WEBSITE_CATEGORIZATION_API_KEY": "your_api_key_here"
}
}
}
}{
"mcpServers": {
"website-categorization-db": {
"command": "python3",
"args": [
"/home/YOUR_USERNAME/mcp-servers/database-lookup/server.py"
],
"env": {
"WEBSITE_CATEGORIZATION_API_KEY": "your_api_key_here"
}
}
}
}
Replace YOUR_USERNAME with your actual system username and your_api_key_here with the API key from Step 2. If you are using a virtual environment, replace python3 with the full path to the Python binary inside the venv, such as /Users/YOUR_USERNAME/mcp-servers/database-lookup/venv/bin/python3.
Claude Code, the CLI version of Claude for software development, supports MCP servers through its own configuration system. You can add a server from the command line using the claude mcp add command, or by editing the .mcp.json file in your project root. Both methods achieve the same result.
# Add the MCP server via the CLI
claude mcp add website-categorization-db \
python3 /path/to/mcp-servers/database-lookup/server.py \
-e WEBSITE_CATEGORIZATION_API_KEY=your_api_key_here{
"mcpServers": {
"website-categorization-db": {
"command": "python3",
"args": ["/path/to/mcp-servers/database-lookup/server.py"],
"env": {
"WEBSITE_CATEGORIZATION_API_KEY": "your_api_key_here"
}
}
}
}
Cursor, the AI-powered code editor, also supports MCP servers. Open Cursor and navigate to Settings, then find the MCP Servers section. Click to add a new server and fill in the configuration details. The command should be python3 (or the full path to your venv Python), with the server script path as the argument. Set the environment variable for your API key in the environment configuration field.
// In Cursor: Settings → MCP Servers → Add Server
Name: website-categorization-db
Command: python3
Args: /path/to/mcp-servers/database-lookup/server.py
Env: WEBSITE_CATEGORIZATION_API_KEY=your_api_key_hereAfter configuring your AI assistant, restart it to load the new MCP server. You should see the Database Lookup tools appear in the available tools list. Try the following example prompts to verify everything is working correctly:
# Test single domain lookup
"What category is bbc.co.uk?"
# Test bulk lookup
"Categorize these domains: amazon.com, etsy.com, shopify.com,
walmart.com, and target.com"
# Test search
"Find technology domains from Japan in the database."
# Test database stats
"Show me the current database statistics."
# Test credit check
"How many API credits do I have remaining?"If the AI assistant responds with categorization data, the server is working correctly. If you encounter errors, double-check that your API key is set correctly, the file path to the server script is accurate, and all dependencies are installed.
The combination of instant lookups, bulk processing capability, and rich categorization data makes this server suitable for a wide range of professional and technical applications.
Analyze large lists of domains quickly by feeding them through the bulk lookup tool. Researchers, analysts, and data teams can categorize thousands of domains in minutes by batching them into groups of 100, creating categorized datasets for further analysis without waiting for real-time classification of each individual site.
Enrich bid request data in programmatic advertising pipelines with IAB category information and brand safety signals. Because database lookups return in milliseconds, they can be integrated into real-time bidding workflows where latency budgets are tight and decisions need to be made in under 100 milliseconds per impression opportunity.
Build parental controls, corporate web filters, and safe browsing solutions using the web filtering category data. Each domain is classified into categories like adult content, gambling, social media, news, and more, enabling you to block or allow domains based on content policy rules without inspecting the actual page content in real time.
Enrich prospect lists and CRM data with domain categorization, buyer persona information, and authority scores. Sales teams and marketing platforms can automatically classify prospect websites to understand their industry, target audience, and web authority, enabling better lead scoring, segmentation, and personalized outreach strategies.
Domain investors, registrars, and marketplace operators can categorize entire portfolios of domains to understand distribution across industries, identify undervalued domains by category, and create filtered browsing experiences for buyers. The bulk lookup tool makes it practical to classify portfolios of thousands of domains efficiently.
Financial institutions, payment processors, and regulated businesses can screen merchant and partner websites against compliance policies. The web filtering categories and content classifications help identify websites involved in gambling, adult content, pharmaceuticals, or other regulated industries that may require additional review or restrictions.
We offer two MCP servers with different strengths. Choose the right one for your workflow, or use both together for maximum flexibility.
| Feature | Database Lookup | Real-Time API |
|---|---|---|
| Response Speed | Instant (milliseconds) | 2 to 10 seconds per URL |
| Coverage | 30M+ pre-classified domains | Any URL on the internet |
| URL Depth | Domain-level only (e.g. stripe.com) | Full subpage support (e.g. bbc.co.uk/news/tech) |
| Data Freshness | Quarterly database updates | Real-time analysis of live page content |
| Taxonomy Support | IAB v2/v3 + Web Filtering | IAB v2/v3, IPTC, Web Filtering, and more |
| Enrichment Data | Personas, PageRank, Country | Full enriched data including technologies, quality scores |
| Bulk Operations | Up to 100 domains per batch | One URL per request |
| Search by Category | Yes, reverse lookups supported | Not available |
| Best For | Speed-critical and bulk analysis workflows | Subpage classification and fresh content analysis |
Many teams use both servers together. The Database Lookup server handles the fast, high-volume domain-level work, while the Real-Time API server handles subpage classification, newly registered domains, and cases where the database returns no result. You can install both MCP servers simultaneously in Claude Desktop, Claude Code, or Cursor.
The database contains over 30 million pre-classified domains, representing the vast majority of active websites on the internet by traffic volume.
Our database is built by continuously crawling and classifying the most visited websites globally, along with a comprehensive sweep of the long-tail web. The result is a collection of over 30 million domain records, each enriched with multiple layers of categorization data. In terms of real-world traffic, these 30 million domains account for an estimated 99 percent or more of all HTTP requests on the internet, because the distribution of web traffic follows a power law where the top millions of domains receive the overwhelming majority of visits.
The database includes major global websites, regional and country-specific sites, niche industry portals, e-commerce storefronts, news publications, blogs, corporate sites, SaaS platforms, government domains, educational institutions, and virtually every other category of web property. Below are example records for a few well-known domains to illustrate the type and depth of data returned by the lookup tools.
Note: The example data above is illustrative. Actual data fields and values may vary. The database is updated quarterly to ensure domain classifications stay accurate as websites evolve and new domains are added.
Every domain lookup through the Database MCP server costs one credit. If you look up a single domain with lookup_domain, that is one credit. If you run a bulk_lookup of 100 domains, that is 100 credits. The database_stats and check_credits tools do not consume credits. This pricing model is identical whether you use the MCP server or the API directly.
New accounts receive free trial credits to get started. Visit the pricing page for full details on credit packages, volume discounts, and enterprise plans.
View Pricing PlansCommon questions about the Database Lookup MCP server, its capabilities, and how it compares to other options.
bbc.co.uk, stripe.com, or nytimes.com, and returns the stored categorization data for that domain. It does not support individual subpages or URLs with paths. If you need to classify specific subpages, such as a particular article or product page, use the Real-Time API MCP server instead. The Real-Time server crawls and classifies the actual content of any URL on the internet, including deeply nested subpages.
bulk_lookup tool accepts up to 100 domains per request. Each domain in the batch counts as one credit. If you need to process more than 100 domains, you can make multiple batch requests sequentially. For example, to categorize 500 domains, you would make 5 batch requests of 100 domains each. The AI assistant can automate this for you by splitting a larger list into batches and processing them in sequence, then combining the results into a single response.
Create your free account, grab your API key, and connect the Database Lookup MCP server to your AI assistant. You can be running domain categorization queries in under five minutes.