MCP Server

The MCP server for B2B data

Add one URL to Claude, Cursor or VS Code and sign in. Your assistant can search and enrich people and companies, find verified work emails and search job postings, with the BlitzAPI docs built in.

WORKFLOWMCP SERVER
LIVEBLITZ API · V2

What your AI gets

One URL gives your assistant three things. The docs and the skills load when it needs them, not up front, so they stay out of its context until then.

API tools

One tool for every BlitzAPI endpoint. Your AI searches, enriches and fetches the data itself instead of only writing code, and each tool takes the same parameters as its endpoint.

Docs access

Your AI searches and reads the full documentation (endpoints, parameters, response schemas, guides) before it calls anything, instead of guessing a parameter name.

Built-in skills

BlitzAPI's own playbooks, served as MCP resources, teach your AI how to use the server and build GTM workflows, like handing a large job to a script.

Setup

Connected in one minute

Add the URL to your AI tool, then sign in with your BlitzAPI account in the window it opens: there is no key to copy. A client that can't sign in sends your API key in the x-api-key header instead.

MCPhttps://api.blitz-api.ai/mcp
# In a new Claude chat, click +,
# then Manage connectors and
# Add custom connector.
# Remote MCP server URL:
https://api.blitz-api.ai/mcp
# Click Connect and sign in.
claude mcp add --transport http blitz-api https://api.blitz-api.ai/mcp
# Then run /mcp in Claude Code, select blitz-api and sign in.

# Prefer an API key? Pass it as a header instead:
claude mcp add --transport http blitz-api https://api.blitz-api.ai/mcp \
  --header "x-api-key: YOUR_API_KEY"
{
  "mcpServers": {
    "blitz-api": {
      "url": "https://api.blitz-api.ai/mcp"
    }
  }
}
{
  "servers": {
    "blitz-api": {
      "type": "http",
      "url": "https://api.blitz-api.ai/mcp"
    }
  }
}
{
  "mcpServers": {
    "blitz-api": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "https://api.blitz-api.ai/mcp"]
    }
  }
}

Five prompts to try

Paste one into your AI tool once the server is connected. The last one runs a large job the right way: as a script, not as one giant tool call.

  1. Find a decision maker

    Using BlitzAPI, find the Head of Sales at stripe.com and get their work email.

  2. Build a prospect list

    Find 25 VPs of Marketing at B2B SaaS companies in the US with 50 to 200 employees. Show name, title, company and profile URL in a table.

  3. Check your account

    How many records do I have left on my BlitzAPI plan, and what's my rate limit per endpoint?

  4. Ask the docs

    What parameters does the waterfall ICP keyword search accept? Show me a complete example JSON body.

  5. Write the pipeline

    Write a Python script that takes a list of company domains, converts each to a LinkedIn URL, then finds the VP of Sales using BlitzAPI waterfall search. Save the results to a CSV and include retry logic.

Large jobs

Built for agents that run at scale

Tool results come back into your AI's context, so big pulls get expensive fast. The server's guidance steers your assistant toward what holds at volume.

A script, not a giant tool call

Beyond a small lookup, your AI writes a Python or JavaScript script that fetches the data and saves it locally, keeping big payloads out of the chat.

Cursors and depth ceilings

Search tools page with max_results and a cursor. A search stops at 50,000 results for people, company and TAM searches, 10,000 for the employee finder and 5,000 for jobs: slice a bigger pull into narrower filters.

Errors it can fix

A 422 names every wrong field by its full path and suggests the closest valid field or value, so your assistant corrects its own call.

SEARCH RESPONSE200
RESULTS_LENGTH2MAX_RESULTS10cursor"eyJzIjpb…"
CURSOR · DEPTH CEILING

One tool per endpoint

Tools are named after what they return, and each API tool runs exactly one endpoint: what the API reference says about a parameter or a response applies unchanged.

Tools
24
API endpoints
20

People

Companies

Jobs

Docs

  • docs_search Search the docs A question in, matching documentation snippets with links out. Where your AI starts.
  • docs_read Read a docs page A documentation page in, its full text out.

Account and utilities

  • util_changelog Changelog The API changelog, newest first, so your AI knows what changed. Uses no records.
  • send_feedback Feedback Sends a bug report, a feature request or feedback to the BlitzAPI team.

Records

Tool calls run on your BlitzAPI plan and count exactly like API calls. One record is one item the API returns to you: one company, one person, one email address, one phone number or one job posting. A call that finds nothing counts nothing.

See plans

Where to go next

The same data from your code, your workflows or an agent.

MCP SERVER

Your AI tool calls the same endpoints as your code, with the same parameters and the same responses.

API Docs
SOURCERESULTS

Frequently asked questions

What is the BlitzAPI MCP server?

A remote MCP server at https://api.blitz-api.ai/mcp that connects Claude, Cursor, VS Code or any MCP client to BlitzAPI. Your AI gets one tool per API endpoint to search and enrich people and companies, find work emails and US phone numbers and search job postings, plus the documentation and built-in GTM skills.

Which AI tools does it work with?

Any client that supports remote MCP servers. Claude on the web, Claude Code, Cursor and VS Code connect with the URL alone. Older versions of Claude Desktop don't support remote servers natively and connect through the mcp-remote wrapper.

How do I sign in: account or API key?

Sign in with your BlitzAPI account: your AI tool opens a browser window, and there is no key to copy. A client that can't sign in can send your API key in the x-api-key header. Never put the key in an Authorization: Bearer header, which only accepts OAuth access tokens, and never commit it to a repository.

Does the MCP server use my records?

Yes. Tool calls run on your BlitzAPI plan and count records exactly like the matching API calls: what a tool returns counts, and a call that finds nothing counts nothing. Checking your plan or reading the changelog uses no records.

Is there a docs-only MCP server?

Yes. https://docs.blitz-api.ai/mcp gives your AI tool the documentation only, with no account and no records used. The main server, https://api.blitz-api.ai/mcp, already includes the docs through its docs tools, so you don't need both.

Can an AI agent buy BlitzAPI data without an account?

Yes, through the BlitzAPI agent storefront, run with ZeroClick: an agent reads the catalog, pays from a prepaid balance and calls the API in the same run, on its own or with its human's approval. The MCP server is for people and teams who work with a BlitzAPI account.

Data shouldn't be
the problem

Grab. Connect. Build.