# Lusha Model Context Protocol (MCP) Server

> **Audience:** Admins and developers setting up the Lusha MCP Server for the first time, or evaluating what the integration can do at a technical level. If you're already connected and want to know how to use Lusha inside Claude, ChatGPT, or Codex, see [Using Lusha Inside LLMs](/user-guide/mcp/using-lusha-inside-llms).


The Lusha MCP Server is a [Model Context Protocol](https://modelcontextprotocol.io/) server that exposes Lusha's B2B contact and company data as callable tools. MCP-compatible AI clients — including Claude, ChatGPT, Codex, Cursor, and VS Code — can invoke those tools through natural language, without any custom API integration. Lusha is also listed on [Gong Collective](https://collective.gong.io/integrations/lusha-systems-inc), so Gong customers can connect the same MCP tools straight into Gong's AI Briefer.

## What the MCP Server exposes

Once connected, AI clients can call the following Lusha capabilities as tools:

- **Contact and company search** — find prospects by name, title, industry, location, and more
- **Profile enrichment** — retrieve verified emails and phone numbers for a contact or company
- **Prospect table building** — assemble targeted lists inside the AI conversation
- **AI Recommendations** — pull ICP-tailored contact and company suggestions directly into the conversation
- **Signal Score** — rank any contact or company list on a 0–100 buying-signal scale using the `signal_score_contacts` and `signal_score_companies` tools; see [Signal Score and ICP Fit Score](/user-guide/ai-recommendations/signal-score-and-icp-fit-score) for how scores are calculated
- **Lookalikes** — find companies and contacts similar to your best customers, then enrich them in the same flow
- **Website Visitors** — identify which companies are browsing your site and act on them without leaving the conversation
- **Conversation Intelligence** — search your account's recorded sales calls by date, company, participant, or keyword, then pull a call's summary, coaching scores, chapters, or transcript, using the `conversations_search` and `conversations_transcript_get` tools; see [Conversation Intelligence API](/user-guide/lushas-api/conversation-intelligence-api) for credit costs and what's included


## Requirements

- A Lusha account on any plan — every user can connect their own Lusha MCP Server, subject to your account's MCP access permission. MCP connects over OAuth, not an API key, so it isn't gated by Admin or Manager role, or by your account's API key permission settings. Admins can still restrict MCP separately: a dedicated **MCP access** permission in **Account Settings → Permissions** is on by default — turn it off and only Admins and Managers can connect.
- Available API credits — MCP calls consume credits from your account's credit pool
- An MCP-compatible AI client (Claude, ChatGPT, Codex, Cursor, VS Code, Gong AI Briefer, or any client that supports the MCP protocol)


Your personal MCP Server URL is in [API Hub → MCP](https://dashboard.lusha.com/api/mcp).

## Connect Lusha now

[Connect via Claude](https://claude.ai/directory/connectors/lusha) — click **Add to Claude** and complete OAuth authentication.

[Connect via ChatGPT](https://chatgpt.com/apps?q=lusha) — find Lusha in the app directory and click **Connect**.

[Connect via Codex](https://docs.lusha.com/mcp-docs) — find Lusha in the Codex plugin directory and click **Connect**.

[Connect via Gong](https://collective.gong.io/integrations/lusha-systems-inc) — in Gong, go to **Settings → MCP connections**, select Lusha, authenticate, and enable the tools you want for AI Briefer.

blockquote
💡 **Starter prompt included:** Every Lusha MCP connection ships with a ready-to-use starter prompt on the consent screen — copy it in one click to run a real example (find a person, verify they still work there, and pull their title and verified contact details) the moment you connect. This applies to all MCP clients, not just Claude.

## Full setup guide and documentation

For complete setup instructions, available tool schemas, example prompts, troubleshooting, and security guidance, visit the full MCP documentation on the Lusha developer portal:

blockquote
👉 [MCP Server documentation on docs.lusha.com](https://docs.lusha.com/mcp-docs)

## Legal notices

(a) Data brokers and other third-party platforms may not embed, expose, or otherwise provide access to the Lusha API, or to Data obtained through it, on or through their own website, product, or service, without Lusha's prior written consent.

(b) Data obtained through the API may not be used to train or develop AI/ML systems, subject to the terms and exceptions outlined in the [Terms and Conditions](https://lusha.com/legal/terms).