# Connect AI Apps And Other Tools: API And MCP > Connect any AI app that supports MCP (Claude, ChatGPT, Gemini, Perplexity, Grok, Cursor and more), or any other app through the Behavly API, with your role and your plan's limits. Source: https://behavly.io/help-center/agent/connect-ai-apps-mcp · Behavly Help Center · Behavly Agent Behavly can be used from outside the app in two ways. AI apps such as Claude, ChatGPT, Gemini, Perplexity, Grok and Cursor connect to Behavly's **MCP server** and work in your workspace for you. Any other app or script uses the **Behavly API**. Both run the same actions as Behavly Agent: campaigns, Agents, contacts and patterns, the Unibox, templates, insights, email accounts and settings. #### Connect An AI App Any AI app that supports MCP can connect: Claude, ChatGPT, Gemini, Perplexity, Grok, Cursor, Claude Code, VS Code and more. The steps are the same everywhere; only the app's menus differ, and **Settings → Account & Settings → MCP** shows them for each app. 1. Copy the server URL from **Settings → Account & Settings → MCP**. 2. In your AI app, open its connector or MCP server settings and add a new server (some apps call it a **custom connector**). Name it **Behavly** and paste the URL. 3. Connect. Behavly's sign-in page opens. 4. Choose **Continue with Behavly** and pick the workspace and the access, or paste a Behavly API key instead. 5. Turn **Behavly** on in a conversation and ask for what you need. - **Claude**: In claude.ai or the Claude app, go to **Customize → Connectors** and click **Add custom connector**. - **ChatGPT**: Turn on **Developer mode** under **Settings → Apps & Connectors → Advanced settings**, then click **Create** and choose OAuth. Developer mode needs a paid ChatGPT plan. - **Gemini**: In **gemini.google.com**, open **Settings → Connected Apps**, click **Add a custom app** under Custom apps and paste the URL, then sign in. Custom apps need a personal Google account in the US. - **Perplexity**: Open **Account settings → Connectors**, click **+ Custom connector**, choose **Remote**, name it Behavly, paste the URL and choose **OAuth**, then sign in. Custom connectors need a Pro, Max or Enterprise plan. - **Grok**: Open **grok.com/connectors**, click **New Connector**, choose **Custom** and paste the URL. Grok Bot adds it when you ask in chat: say to add a custom MCP server called behavly at the URL, click **Add it**, then **Authorize**. Custom connectors need a paid Grok plan. - **Claude Code**: Run `claude mcp add --transport http behavly https://behavly.io/api/mcp`, then type `/mcp` and choose Behavly to sign in. - **Cursor**: Click **Add to Cursor** in the MCP tab and choose **Install**, or open **Cursor Settings → Tools & MCP**, click **New MCP Server** and add behavly with the URL. Sign in when Cursor asks. - **VS Code**: Run **MCP: Add Server** from the Command Palette, choose **HTTP**, paste the URL and name it behavly. Click **Allow** and sign in when VS Code asks. #### Two Ways To Sign In - **Continue with Behavly**: The recommended way. Sign in, pick the workspace and choose **Read & write** or **Read only**. Behavly makes a key for the app, so there is nothing to copy or paste. The app appears under **Connected apps** on the MCP tab. - **Use an API key instead**: Handy for developers and automations. Paste a key from the API tab: Behavly checks it and makes the app a key of its own with the same access, so your key stays private and the app appears under **Connected apps**. Apps that cannot sign in can send a key as the header `Authorization: Bearer YOUR_API_KEY` instead. Those use that key itself: they show under **Connected apps** as using it, and revoking the key cuts them off, along with anything else using it. #### API Keys 1. Open **Settings → Account & Settings → API**. 2. Click **Create API key**, name it after the app or computer it is for, and choose **Read & write** or **Read only**. 3. Copy the key. It is shown once: Behavly keeps only a fingerprint of it, so a lost key is revoked and replaced, never recovered. The [API documentation](https://behavly.io/api/v1/docs) lists every endpoint with examples. Every call is `POST /api/v1/{action}` with a JSON body and the key in the `Authorization` header. #### What A Key Can Do - **One person, one workspace**: A key acts as the person who made it, in the workspace it was made in. It never reaches another workspace. - **Their role, as it is now**: A key can do what its person's role allows at the moment it is used. Change someone's role and their keys follow; remove them from the workspace and their keys stop. - **Read only**: Looks things up. Every change is refused. - **Deleting and sending**: By default, deleting something or sending email to real people waits for you: the app gets a link, and you check the change in Behavly and approve or decline it within 30 minutes. #### Plan Limits Apply Every plan can use the API and MCP. Everything an app does counts toward your workspace's plan exactly as if you did it in Behavly: AI-written emails, AI email designs, reply drafts, contacts, mailboxes and active Agents. When a limit is reached, the app is told what ran out and given the link to **Settings → Billing**. #### Choose Areas Behavly offers about 150 tools to an AI app. Some apps load only so many (Cursor about 40, VS Code 128). End the server URL with the areas you use there, for example `https://behavly.io/api/mcp?areas=campaigns,inbox`. The areas are `campaigns`, `agents`, `contacts`, `inbox`, `templates`, `insights`, `mailboxes` and `settings`. #### Revoke A Key Or Turn Access Off - Revoke a key from its row in the API tab. Disconnect an app from **Connected apps** on the MCP tab. Either way it stops working right away. - Owners and admins see every key in the workspace, with who made it, and can revoke any of them. - **Workspace access** in the API tab turns every key off at once without deleting any. - With AI switched off in **Settings → Account & Settings → AI & Privacy**, AI apps cannot connect at all. - Changes made from outside Behavly appear in the [Audit log](https://behavly.io/help-center/privacy/audit-logs) as **via MCP** or **via API** with the key's name. See also: https://behavly.io/help-center/privacy/ai-controls, https://behavly.io/help-center/privacy/audit-logs, https://behavly.io/help-center/plans/what-counts-against-a-limit