MCPs and APIs for Lead Generation 101 Guide
With AI being increasingly used for go-to-market, the letters “MCP” have found their way into more LinkedIn, Reddit, and Substack feeds when it comes to lead generation. But why, and should you even care?
While it sounds like a term that firmly belongs in the developer’s world, if you run your own outreach, it’s time to pay attention.
Prospecting is changing, and MCP is the key to connecting your Hunter go-to-market data into chat-based prospecting in Claude, ChatGPT, and Perplexity.
The short story:
- An MCP (Model Context Protocol) connection is the standard plug that lets an AI assistant like Claude, ChatGPT, or Perplexity use your Hunter account from a conversation. You ask in plain English, and the assistant does the Hunter work for you, meaning you don’t need to use the Hunter app interface..
- An API is the counter where one piece of software asks another for something specific. Hunter's APIs answer questions like "who works at this company?" and "is this email address safe to send to?"
- The MCP sits in front of Hunter's full API, meaning one connection gives your AI assistant 30+ tools to identify companies, find and verify email addresses, enrich contacts, and manage your leads and email sequences.
- The biggest benefit is that it prevents an AI assistant from hallucinating data. With an MCP connection, every contact comes back found and verified, or flagged as something Hunter couldn't confirm, so you never send to an invented email address.
- The MCP is included in every Hunter plan. Follow these quick start instructions:
- Connect Hunter to Claude by adding the .md file inside a Claude chat
- Connect Hunter to the ChatGPT app by adding the .md file inside a ChatGPT chat
What is an MCP?
MCP stands for Model Context Protocol. It's an open standard, published by Anthropic (the company behind Claude) in 2024, that gives AI assistants one shared way to connect to outside tools and data.
The comparison often used is the wall socket.
Before wall sockets were standardized, every appliance needed its own wiring job. Now any plug fits any socket, and nobody thinks about it. MCP does the same for AI assistants: any tool that speaks the protocol can connect to any assistant that speaks it, without a custom integration built for each pairing.
This is important because while AI assistants are smart, they lack access to your specific data.
Claude can describe a great-fit customer with ease. But, it cannot identify the companies and verified email addresses of decision makers inside.
AI assistants’ knowledge stops on the data it is trained on. That’s why you need to connect your hunter.io account as a key to unlock your lead generation.
Once Hunter’s MCPis plugged in, your AI assistant can search companies, find contacts, verify addresses, and save leads using your actual account, with your plan and your credits, from inside the conversation you're already having.
You don't need to understand how the protocol works underneath, any more than you need to understand the wiring behind a socket.
You set it up once and forget it exists.
What is an API?
An API (Application Programming Interface) is the counter where one piece of software asks another for something specific. It's a published menu of requests the service will answer, each with a defined question and a defined answer.
Hunter's Email Verifier API, for example, takes one email address and returns a verdict: valid or not, with a confidence score and the detail behind it.
The Domain Search API takes a company's website and returns the email addresses Hunter has for the people who work there, with sources attached.
APIs are precise, fast, and reliable, which is exactly why products get built on them.
Unfortunately, APIs have long sat in the world of developers, with the ability to write code an immediate barrier to those less-technically minded.
The MCP eliminates that barrier.
How does an API and MCP work together?
If the Hunter API is the service counter, the MCP server is the translator standing at it on your behalf.
Breaking down the MCP and API
You type something like "find the marketing decision-makers at these 20 companies and verify their email addresses" into Claude.
Claude decides which tools to use, whether that's a Domain Search, an Email Verifier check, or something else entirely, instead of you needing to think about the right solution.
The MCP server takes each request and places it at Hunter's API counter in the format the API expects.
It then collects the responses and hands them back to the assistant, which turns the raw data into the table you actually wanted.
The MCP gives you the access without tech knowledge
The API does the work it has always done, and the MCP adds the translation in front, letting you place orders in plain English instead of code.
One connection covers the whole counter, because Hunter's MCP exposes the full Hunter API as over 30 individual tools, and the AI assistant picks the right ones for whatever you ask.
Staying in control of your Hunter credit usage
Critically, you remain in control of the credits you use.
In Claude, each tool allows you to set a permission level, so you can let read-only lookups run freely while anything that changes your account or uses credits waits for your approval.
Verifying an email costs 0.5 credits and saving a lead costs 1 credit, and with the recommended settings, neither happens without you saying yes.
What APIs are available from Hunter?
Since the MCP is a window onto the API, it helps to know what's actually on the menu.
Hunter's API breaks into groups.
Find companies and people
Discover API
The Discover API finds companies matching your criteria, whether you describe them in natural language or with filters for industry, size, location, and technology, and it's free to call.
The Email Count endpoint tells you how many addresses Hunter has on file for a company before you spend anything.
Domain Search API
The Domain Search API returns the email addresses found for a company, with names, roles, and the public sources each one came from (1 credit per 1-10 addresses).
Domain Search API
The Email Finder API takes a person's name and their company's website and returns their most likely professional address with a confidence score (1 credit).
Verify and enrich data
Email Verifier API
The Email Verifier API checks if an address is deliverable and returns a status and score, at 0.5 credits per check. It ensures you find verified and accept-all data to improve your bounce rate.
Lead Enrichment API
Lead Enrichment turns an email address into a person: name, role, location, and social profiles.
Company Enrichment API
Company Enrichment turns a website into company detail: industry, size, description, and the technology it runs on.
Combined Enrichment API
Combined Enrichment does both from a single address. Each costs 0.2 credits, and only when the core data actually comes back.
If you used Clearbit for enrichment before it was folded into HubSpot, Hunter's API works as a direct replacement, down to a Clearbit-format response option.
Logo API
The free Logo API returns any company's logo from its website address, ready to be used in your CRM, landing pages, dashboards, or product.
Manage leads and send emails
Leads API
The Leads API and its neighbors (Lead Lists, Custom Attributes) let you save, organize, and update the contacts you've found.
This enables you to segment your contacts and companies based on criteria like location, technology, owner, or whatever attribute you choose.
Sequences API
The Sequences API adds recipients to your outreach sequences and starts them.
API key basics
If you want to go deeper, the full API documentation covers every endpoint with examples.
Your API key lives at hunter.io/api-keys, and it should be treated the same as any password, meaning you never share it or paste it anywhere public.
An Account endpoint reports your plan, credits, and usage, free of charge, which is how your AI assistant can tell you whether a big job fits your remaining credits before it starts.
For experimenting without spending credits, the special key test-api-key returns dummy responses on the Domain Search, Email Finder, and Email Verifier endpoints.
If you connect through Claude's connector settings or the ChatGPT app, you may never touch a key at all, because the connection uses OAuth: you log in on Hunter's own page, and the assistant never sees or stores your password.
How is the Hunter MCP being used?
If you're handling go-to-market by yourself, the MCP can be the foundation for an entire Claude prospecting workflow.
Claude can act as a research agent, identifying great-fit companies, finding the contacts with decision-making power, and then using Hunter to get their verified email addresses.
Users work this way because it blends high-quality, trustworthy data with less context switching and browser hopping, the kind that wastes time and pulls you away from other work.
The AI assistant does the reasoning, Hunter gives it the data you can trust, and you can work on something else all the while you’re running your lead generation.
Here's what that can look like in each assistant.
If you're handling go-to-market by yourself, the MCP can be the foundation for an entire Claude prospecting workflow.
Claude can act as a research agent, identifying great-fit companies, finding the contacts with decision-making power, and then using Hunter to get their verified email addresses.
Users work this way because it blends high-quality, trustworthy data with less context switching and browser hopping, the kind that wastes time and pulls you away from other work.
The AI assistant does the reasoning, Hunter gives it the data you can trust, and you can work on something else all the while you’re running your lead generation.
Here's what that can look like in each assistant.
Claude
Claude's strength is reasoning over everything at once.
Paste in the websites of customers you've already closed and ask it to work out the profile they share, then have Hunter find companies matching that pattern.
Ask it to score every match against the profile with a reason per score.
Then, have Hunter turn the top companies into named decision-makers with verified addresses, and finish with a one-line outreach hook per contact.
Claude can even build the results into an interactive account-scoring sheet you can sort, edit, and reopen later.
The connector works on Claude's Free, Pro, and Max plans, and on the Free plan Hunter can be the one custom connector you're allowed.
Click here to learn more about Claude x Hunter.
ChatGPT
The Hunter ChatGPT app suits starting from a customer you already win with.
Give it one website and ask for lookalike companies, and Hunter Discover returns a ranked list with the number of addresses on file for each.
Narrow it in a sentence ("just the European ones"), ask which recently raised money or are hiring, then run a Domain Search to get named contacts with verification statuses, and have it draft a three-sentence opener for the first one.
The process runs on ChatGPT's free plan; you call the app by typing @ and then Hunter.
Click here to learn more about ChatGPT x Hunter.
Perplexity
Perplexity reads the open web as it is today and cites where each claim came from, so it's great to use when you're building a lead list from start, for instance with companies that raised this quarter.
It builds the market map with sources attached, reasons about which companies fit what you sell, and then hands the qualified list to Hunter for named contacts and verified addresses.
One thing to know upfront is that custom connectors need a paid Perplexity plan, Pro or higher, so the free tier can't add Hunter yet.
Click here to use Hunter in Perplexity.
Prompts to test your MCP
Whatever your AI assistant, once set up, try these prompts:
- "Before we start, pull my Hunter account details: the plan I'm on, what I've used this month, and what's left. Then estimate what this job will use and tell me whether it fits."
- "These are deals I lost a year ago and want another run at. Verify each contact's email with Hunter. Where an address comes back invalid, the person has probably moved on, so run a Domain Search on the company and find whoever now holds the role."
- "Pull the stats for every sequence I sent this quarter and rank them by reply rate. Name the winner and the loser, and tell me what separated them."
Why use Hunter's MCP?
Broadly, the Hunter MCP will enable you to trust the business data you're using to orchestrate outbound in an AI assistant:
More specifically, the MCP's benefits spread across four areas:
1) Data hallucinations
Ask an AI assistant for someone's email address and it will often produce one that looks right and doesn't exist, because guessing plausible text is what these models do.
Send to enough of those and your bounce rate climbs, and past 2% your sender reputation starts paying for it.
With the MCP connected, the assistant's answers come from Hunter's records instead of its imagination.
Addresses arrive with a verification status and confidence score, accept-all servers get flagged as the warning they are, and where Hunter has nothing, it says "no data" rather than inventing a name.
Knowing exactly which contacts you can send to today, and which need more work first, is the whole game in deliverability.
2) Limit context switching
The old productivity loop was research in one window, Hunter in another, a spreadsheet in a third, and CSV exports moving the same names between them.
Contact data goes stale quickly (22% of email data decays each year), so every manual hop is also a chance for the list to age before you use it.
With the MCP in place, a contact moves from research to a verified leads list to a sequence inside one conversation, and the re-verify step is a sentence, not an export.
3) No added cost
The MCP connection costs nothing extra. It's included in every Hunter plan, your existing credits and limits apply, and a free Hunter account works from day one.
4) You're in control
Unlike AI assistants, whose token usage is something you have to actively control by selecting a different model, reasoning level, and/or choose to use the assistant in a browser/desktop app, the Hunter MCP gives you full control over Hunter credit usage.
It's built to keep you in control, with OAuth sign-in and per-tool permissions, so the assistant asks before it does anything that spends credits or changes your account.
Go beyond the MCP
The MCP is the newest door into Hunter's API, but far from the only one. In fact, Hunter’s API is being used to grow businesses across outreach, products, and marketing experiences.
Build products
Vocations, a job-signal platform for recruiters, uses the Email Finder API to power the contact feature at the heart of its product. The integration went live in a day, and today its 300 daily users find around 1,500 decision-makers through it.
Make verification part of the prospecting routine
Modgility, a RevOps consultancy with a two-person sales team, builds lists of companies running HubSpot, verifies the contacts, and runs targeted sequences. That outbound accounts for 25% of their new business.
Feed your CRM
WebFootprint, a software agency selling to fewer than 200 possible buyers nationwide, pipes Hunter's data into a custom CRM, where it fills every company and contact record and surfaces new hires at target accounts automatically.
The founder puts the savings at an hour a day, and he's been a Hunter customer since 2016.
Those are very different businesses placing orders at the same counter.
The MCP simply means you no longer need an engineering background to place one yourself.
Use the Hunter MCP
The fastest way to connect Hunter is fitting - let the AI assistant handle its own setup.
We've written step-by-step instructions designed to be pasted directly into a chat.
The assistant reads them, greets you, and walks you through the whole connection one confirmation at a time, from checking you have a Hunter account through to running your first real prompt.
You won't have to figure anything out on your own.
- Claude: Copy the Hunter for Claude setup instructions here into a new Claude chat.
- ChatGPT: Copy the Hunter for ChatGPT setup instructions here into a new ChatGPT chat.
- Perplexity: On a Pro plan or higher, add Hunter as a custom connector, then run your first market map.
The setup takes five minutes, and the next time you need a list built, checked, and saved, you'll do it without opening a single extra tab.
Learn more about the MCP
To learn more about the MCP, APIs, and Hunter in your AI assistant, visit our help center content below:
- How to connect Hunter to Claude with MCP, the full walkthrough with screenshots, including the API-key route for Claude Team members without admin rights.
- The Hunter API, every endpoint with credit costs and rate limits.
- Can I use a test API key? for experimenting without spending credits.
- How to verify emails via API, with code examples, if you're ready to write a little code.
- Which plan is right for me? on choosing between the All-in-One Outreach plans and the Data Platform plans.
MCP FAQs
Do I need a paid Hunter plan to use the MCP?
No. The connection works with every Hunter plan, including free accounts. Your existing credits and request limits apply, and some actions use credits, like verifying an email (0.5 credits) or saving a lead (1 credit).
Do I need to be a developer?
Not for the MCP. Connecting through Claude's settings or the ChatGPT app is a few clicks plus a Hunter login, and from then on everything happens in plain English. The API itself is the developer route, for building Hunter's data into a product or an internal system.
What's the difference between the Hunter MCP and the Hunter API?
The API is the underlying service: the set of requests Hunter answers, from Domain Search through to Sequences. The MCP puts a translation in front of that API so an AI assistant can use it through conversation. It's the same data and the same credits, reached a different way.
Which AI assistants does it work with?
Claude (web, desktop, and Claude Code), ChatGPT via the Hunter app, Perplexity on its paid plans, and any other assistant that supports MCP connectors.
Is my data secure?
The connection runs through Hunter's official MCP server using OAuth, meaning you log in on Hunter's own page and the assistant never stores your password. The assistant only accesses your Hunter data within the permissions you grant, and you can revoke the connection at any time.
What can the assistant actually do once connected?
The connection covers the full Hunter API in over 30 tools: Discover, Domain Search, the Email Finder, the Email Verifier, person and company enrichment, Leads with lists, tags, and custom attributes, message templates, and Sequences through to their statistics.
Will the assistant spend my credits without asking?
Not if you set it up the recommended way. Hunter suggests leaving read-only tools free to run while setting anything that writes or spends to "needs approval," so every credit-using action waits for your yes.