Apollo.io MCP Server
OfficialClick on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Apollo.io MCP ServerFind sales directors in San Francisco at tech startups."
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Apollo.io MCP Server
The official Apollo.io Model Context Protocol server for prospecting, lead enrichment, sales intelligence, and outreach workflows.
Provider: Official, first-party server built and maintained by Apollo.io.
Bring Apollo.io prospecting, enrichment, and outreach directly into your favorite MCP-compatible AI application — GitHub Copilot, Visual Studio Code, Cursor, Claude Code, Claude Desktop, and Claude Cowork.
Apollo.io: https://www.apollo.io/
Model Context Protocol: https://modelcontextprotocol.io/
License: MIT
Table of Contents
Related MCP server: apollo-io-mcp-server
Overview
The Apollo.io MCP server exposes Apollo.io's go-to-market data and workflows through the Model Context Protocol, so MCP-compatible AI applications can prospect for people and companies, enrich contacts, query sales analytics, and load leads into outreach sequences using natural language.
Hosted and maintained by Apollo.io. The server runs remotely — there is nothing to build, host, or run locally.
Transport: Streamable HTTP.
Endpoint:
https://mcp.apollo.io/mcpAuthentication: Handled through Apollo.io OAuth. You sign in with your Apollo.io account and authorize access from your MCP client; the server then acts with your Apollo.io permissions. See Authentication and Authorization.
This repository also ships an optional Apollo plugin for Claude Code and Cowork that bundles ready-made workflow skills (/apollo:* commands) on top of the same MCP server.
Key Capabilities
The following capabilities are confirmed by the server tools and skills in this repository:
People prospecting — search Apollo's people database by title, seniority, location, and company attributes.
Company prospecting — search organizations by industry, size, location, and keywords.
Contact and account enrichment — match and enrich people and organizations, individually or in bulk.
Outreach sequence workflows — find sequences, and add or remove/stop contacts in a sequence.
Sales analytics — query email, call, meeting, task, opportunity, sequence, and conversation-intelligence metrics and return formatted breakdowns.
Contact management — create Apollo contacts as part of enrichment and sequence-loading workflows.
The remote server may expose additional tools beyond those exercised by the bundled skills. The list above reflects what is confirmed in this repository. See Available Tools and Capabilities.
Prerequisites
An Apollo.io account. Sign in at https://www.apollo.io/.
Apollo.io permissions and credits. Actions run with your account's permissions, and some actions consume Apollo credits (for example, enrichment). See Safety and Responsible Usage. For plan, credit, and permission details, see the Apollo documentation.
An MCP-compatible client that supports the Streamable HTTP transport, such as:
GitHub Copilot in Visual Studio Code
Cursor
Claude Code
Claude Desktop / Claude Cowork
Other MCP clients (see Generic MCP Client Configuration)
No local runtime required. The server is remote and hosted by Apollo.io — you do not need Node.js, Docker, or a local installation to use it. (The optional Claude Code / Cowork plugin installs workflow skills locally through your client's plugin system, but the MCP server itself remains remote.)
Install in Visual Studio Code and GitHub Copilot
The Apollo.io MCP server is published in the official Model Context Protocol Registry, but it is not yet discoverable in the curated GitHub MCP Registry or the VS Code MCP gallery. Until then, add it manually with the configuration below.
Option A — workspace configuration file
Create a .vscode/mcp.json file in your workspace:
{
"servers": {
"apollo": {
"type": "http",
"url": "https://mcp.apollo.io/mcp"
}
}
}Option B — command palette
Open the Command Palette (
Cmd/Ctrl + Shift + P).Run MCP: Add Server.
Choose the HTTP (Streamable HTTP) server type when prompted.
Enter the URL
https://mcp.apollo.io/mcpand name the serverapollo.
Expected authentication and first run
Add the MCP server using either option above.
Start or enable the
apolloserver.Complete Apollo.io authentication in the browser window that opens.
Approve the requested access for your Apollo.io workspace.
Open Copilot Chat in Agent mode.
Confirm that the Apollo tools are listed and available to the agent.
For the current terminology and behavior, see the VS Code documentation on MCP servers: https://code.visualstudio.com/docs/copilot/chat/mcp-servers.
Install in Cursor
Add the remote Streamable HTTP endpoint to your Cursor MCP configuration (for example, ~/.cursor/mcp.json, or the project-level .cursor/mcp.json):
{
"mcpServers": {
"apollo": {
"type": "http",
"url": "https://mcp.apollo.io/mcp"
}
}
}Then:
Open Cursor settings and confirm the
apolloserver is listed under MCP.Authenticate with your Apollo.io account when prompted.
Approve the requested access, then use Apollo tools from the agent.
This repository also includes a Cursor plugin manifest (
.cursor-plugin/plugin.json) that references the same remote endpoint (.mcp.json), so no local package needs to run.
Install in Claude Code
You can connect the server in either of two ways.
Option A — add the remote MCP server directly
claude mcp add --transport http apollo https://mcp.apollo.io/mcpThen run /mcp inside Claude Code, select apollo, and authenticate.
Option B — install the Apollo plugin (adds /apollo:* workflow skills)
Add this plugin's marketplace:
/plugin marketplace add apolloio/apollo-mcp-pluginInstall the plugin:
/plugin install apollo@apollo-plugin-marketplaceRestart Claude Code so the MCP server starts correctly.
The plugin automatically configures the Apollo MCP server — no manual config files. After installing, authenticate with /mcp and run /apollo:* commands (see Available Tools and Capabilities).
Install in Claude Desktop or Claude Cowork
Claude Desktop and Claude Cowork are separate from Claude Code (the CLI). Use the steps below for the desktop-based clients.
Cowork (one-click)
Click to install in a single step:
Then restart Cowork so the MCP server starts correctly.
Claude Desktop
Add the Apollo connector from your Claude settings, then authenticate:
Open Settings and connect the Apollo.io connector (Apollo appears as a remote MCP server).
Complete Apollo.io authentication in your browser.
Approve the requested access.
Confirm Apollo tools are available in a new chat.
Exact menu labels and navigation vary by Claude Desktop version. See Anthropic's documentation for the current steps to add a remote MCP connector.
Generic MCP Client Configuration
For any other MCP-compatible client that supports Streamable HTTP:
{
"name": "apollo",
"type": "streamable-http",
"url": "https://mcp.apollo.io/mcp"
}Exact property names vary between clients. Some clients use
type: "http"(as in the VS Code and Cursor examples above) and some usetype: "streamable-http". Consult your client's MCP documentation for the precise schema.
Authentication and Authorization
The Apollo MCP server uses Apollo.io OAuth. Only confirmed behavior is documented here.
Sign-in flow. When you connect the server, your client opens an Apollo.io sign-in and authorization window. After you approve access, your client can call Apollo tools on your behalf.
Workspace and permissions. The server acts with the authenticated user's Apollo.io permissions. You can only access data and perform actions your Apollo.io account is allowed to.
Reconnecting. In Claude Code you can re-run
/mcp, select apollo, and re-authenticate. In other clients, re-trigger authentication from the client's MCP or connector settings.Revoking access. Remove or disconnect the
apolloserver from your MCP client to stop it from acting on your behalf. To manage authorizations at the account level, use your Apollo.io account settings.
OAuth authorization-server metadata is discoverable at
https://mcp.apollo.io/.well-known/oauth-authorization-server. Beyond that, this README does not assert a specific OAuth specification or profile.
Usage Examples
Natural-language prompts you can try once the server is connected:
Find software engineering leaders at B2B SaaS companies in Germany with 200–1,000 employees.Enrich these business contacts and return their current title, company, and verified business email when available.Show email and call performance by rep for this quarter, sorted by calls made.Add the selected contacts to my Apollo sequence "Q1 Enterprise Outbound" — show a preview before enrolling.If you installed the Claude Code / Cowork plugin, the same workflows are available as slash commands:
/apollo:prospect VP of Engineering at Series B+ SaaS companies in the US, 200-1000 employees/apollo:enrich-lead https://www.linkedin.com/in/example/apollo:sequence-load add 20 VP Sales at SaaS companies to my "Q1 Outbound" sequence/apollo:analytics Show me team call connect rate this quarter by rep
Review before you act. For consequential actions — enrichment (credit-consuming) and adding contacts to a sequence (potentially outbound) — review the exact targets and parameters before confirming the operation.
Available Tools and Capabilities
The remote server provides the underlying Apollo tools; the tools listed below are those confirmed in this repository through the bundled workflow skills. The full, authoritative tool set is provided by the remote server and may evolve independently of this document.
Workflow skills (Claude Code / Cowork plugin)
High-value skills that chain multiple Apollo tools into complete workflows:
Skill | What it does |
| Drop a name, LinkedIn URL, or email and get a full contact card with company context and next actions. |
| Describe your ICP in plain English and get a ranked table of enriched decision-makers. |
| Find leads, enrich them, dedupe, and bulk-add them to an Apollo sequence with a preview before enrollment. |
| Ask any sales performance question and get formatted tables from real Apollo analytics data. |
Underlying Apollo tools referenced by the skills
Tool | Description | Data effect | Outreach / external action |
| Search Apollo's people database by title, seniority, location, and company filters. | Reads | No |
| Search Apollo's organization database by industry, size, location, and keywords. | Reads | No |
| Match and enrich a single person from available identifiers. | Reads / enriches | No |
| Match and enrich multiple people in one call. | Reads / enriches | No |
| Enrich a single organization. | Reads / enriches | No |
| Enrich multiple organizations in one call. | Reads / enriches | No |
| Create a contact in the Apollo workspace. | Changes Apollo data | No |
| Find outreach sequences by name. | Reads | No |
| Add contacts to an outreach sequence. | Changes Apollo data | Yes — may start outbound depending on sequence settings. |
| Remove or stop contacts in an outreach sequence. | Changes Apollo data | Affects active outreach |
| List connected email/sending accounts. | Reads | No |
| Retrieve sales analytics metrics and breakdowns. | Reads | No |
Some actions (such as enrichment) consume Apollo credits. For per-tool credit costs and the authoritative, up-to-date tool reference, see the Apollo documentation.
Model Recommendations
For clients where you can choose a model (for example, Claude Code via /model):
Opus (best quality) — strongest reasoning and the most reliable multi-step tool orchestration. Best for prospecting workflows, ambiguous matches, multi-step chaining, and high-stakes tasks. Tradeoff: higher latency and usage cost.
Sonnet (faster) — good for quick lookups, lightweight enrichment, and rapid iteration. Tradeoff: may need more guidance on complex multi-step workflows.
Safety and Responsible Usage
Review prospecting results before using or exporting them.
Follow applicable privacy, marketing, anti-spam, and data-protection laws in your jurisdiction and your recipients'.
Confirm recipients before adding anyone to an outreach sequence. Depending on your sequence and sending configuration, enrollment may start outbound automatically. Always verify the sequence name, sending account, and enrollment volume before confirming.
Be aware of Apollo credit consumption. Some actions (such as enrichment) consume Apollo credits, and bulk actions scale with volume. This repository's skills are designed to warn and request confirmation before credit-consuming actions. For credit costs, see the Apollo documentation.
Avoid bulk or destructive actions without reviewing their exact scope first.
Respect Apollo account permissions and your organization's policies.
This section is guidance, not legal advice.
Privacy and Data Handling
Requests are sent to the Apollo-hosted MCP endpoint (
https://mcp.apollo.io/mcp).The server interacts with Apollo data under your authenticated account and permissions.
Avoid sending unnecessary sensitive information in prompts.
For Apollo.io's official privacy and security resources:
Privacy Policy: https://www.apollo.io/privacy-policy
Privacy Center: https://www.apollo.io/company/privacy-center
Security: https://www.apollo.io/product/security
Trust Center: https://trust.apollo.io/
Troubleshooting
Symptom | What to check |
Authentication window does not open | Ensure pop-ups are allowed; re-trigger authentication from your client's MCP/connector settings. |
Authentication session expired | Re-authenticate (in Claude Code, run |
Server is not starting | Restart the client after adding/installing; confirm the endpoint URL is exactly |
Tools do not appear | Confirm the server is enabled and authenticated, and (for Copilot) that chat is in Agent mode. |
"Permission denied" on an action | Your Apollo.io account may lack the required permission for that action. |
Credits unavailable | Enrichment and some actions require Apollo credits; confirm your workspace has credits available. |
Client does not support Streamable HTTP | Use a client that supports the Streamable HTTP transport, or the generic config in Generic MCP Client Configuration. |
Invalid JSON in config | Validate your |
Corporate proxy or firewall blocks the endpoint | Ensure |
Official MCP Registry
This server is published and active in the official Model Context Protocol Registry.
Field | Value |
Registry name |
|
Version |
|
Transport | Streamable HTTP |
Endpoint |
|
Publication in the official Model Context Protocol Registry is separate from the curated GitHub MCP Registry and the VS Code MCP gallery. Being active in the official registry does not automatically make the server discoverable in those curated catalogs.
Development and Contribution
This repository contains the Apollo MCP client configuration, the Claude/Cursor plugin manifests, workflow skills, and the registry metadata. The MCP server itself is hosted by Apollo.io.
Repository layout:
Path | Purpose |
| Remote MCP server definition ( |
| Official MCP Registry metadata (name, version, transport, endpoint). |
| Claude Code / Cowork plugin and marketplace manifests. |
| Cursor plugin manifest. |
| Glama MCP directory metadata. |
| Workflow skills: |
| CI that publishes |
Validating server.json
server.json follows the MCP server schema referenced at the top of the file. Publication is automated by the CI workflow (see below), which downloads the official mcp-publisher and runs mcp-publisher publish.
There is no standalone validate command. To check server.json locally before pushing, run a dry-run publish with the official CLI:
mcp-publisher publish --dry-runContributions are welcome via pull requests. Please keep version metadata consistent across server.json, .claude-plugin/plugin.json, and .cursor-plugin/plugin.json.
Versioning and Releases
The server version is declared in
server.json(version: 0.1.1) and mirrored in the plugin manifests (.claude-plugin/plugin.json,.cursor-plugin/plugin.json).On every push to
mainthat changesserver.json, thepublish-mcp.ymlworkflow publishes the new version to the official Model Context Protocol Registry (authenticating via GitHub OIDC).Keep the version in
server.jsonand the plugin manifests aligned before publishing, so the repository and the registry stay in sync.Releases are driven by
server.jsonchanges onmain— the repository does not cut GitHub Releases or version tags per version.
Security
Please do not report security vulnerabilities through public GitHub issues.
Report suspected vulnerabilities privately through the Apollo Trust Center. See SECURITY.md for the full disclosure process and scope.
Support
Need | Where to go |
Apollo product support (data, credits, account, billing) | |
Apollo product documentation | |
MCP integration bugs (this repository) | Open a GitHub issue on apolloio/apollo-mcp-plugin |
Feature requests | Open a GitHub issue on apolloio/apollo-mcp-plugin |
Security vulnerabilities | Do not open a public issue — see Security |
Credits
License
MIT — see LICENSE for details. Copyright (c) 2025 Apollo.io.
This server cannot be installed
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Related MCP Servers
- AlicenseAqualityCmaintenanceMCP server for Apollo.io B2B database enabling people and company search, enrichment, and insights through conversational interfaces.Last updated949119MIT
- FlicenseBqualityDmaintenanceExposes Apollo.io API functionalities as MCP tools for people and organization enrichment, search, and job postings. Enables natural language interaction with Apollo.io data.Last updated516
- FlicenseBqualityBmaintenanceA Model Context Protocol (MCP) server for the Apollo.io API, giving AI coding assistants direct access to Apollo.io's sales intelligence platform for prospecting, enrichment, CRM operations, and outreach.Last updated505
- AlicenseAqualityDmaintenanceMCP server for Apollo.io B2B sales intelligence API. Enables searching and enriching people and organizations via natural language.Last updated7323MIT
Related MCP Connectors
MCP server for LeadDelta — manage LinkedIn connections and CRM data via AI assistants.
Managed LinkedIn MCP server for AI agents: search, connect, message and enrich on accounts you own.
Remote MCP server to enrich company profiles with structured B2B data and confidence scores.
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/apolloio/apollo-mcp-plugin'
If you have feedback or need assistance with the MCP directory API, please join our Discord server