---
title: Using MCPLambda with AI Clients
description: Learn how to connect your MCPLambda deployments to popular AI coding tools and clients.
---

Once your MCP server is deployed on MCPLambda, you can connect it to any AI client that supports the Model Context Protocol. MCPLambda provides a single, high-performance endpoint for your server using **Streamable HTTP** (modern) or **SSE** (legacy), making integration as simple as copying a URL.

---

## 1. Connecting via Deployment URL (Recommended)

Most modern AI clients, like **Cursor** and **VSCode extensions**, allow you to connect directly to remote MCP servers using an HTTP or SSE URL.

### Cursor
Cursor supports MCP servers natively. Using the MCPLambda deployment URL is the most efficient way to give Cursor access to your tools.

1. Copy your **Deployment URL** from the MCPLambda dashboard.
2. Open **Cursor Settings** > **General** > **MCP**.
3. Click **+ Add New MCP Server**.
4. **Name:** Give it a name (e.g., `mcpl-github-server`).
5. **Type:** Select `SSE` (Cursor uses this option for all HTTP-based remote transports).
6. **URL:** Paste your MCPLambda deployment URL.
7. Click **Save**. Cursor will now initialize and list your tools.

### VSCode (via Roo Code / Continue / Cline)
Extensions like Roo Code and Cline allow you to connect to remote MCP servers directly via their settings.

1. Copy your **Deployment URL**.
2. Open the extension's MCP settings.
3. Add a new server with the type **SSE/HTTP**.
4. Paste your URL and save.
5. Your tools will automatically appear in the extension's toolbelt.

---

## 2. Connecting Claude Desktop

For clients that support remote MCP servers via configuration files, you can simply add the deployment URL.

### Claude Desktop
While Claude Desktop primarily uses local `stdio` servers, it also supports remote servers via URL.

1. Open your `claude_desktop_config.json`:
   - **macOS:** `~/Library/Application Support/Claude/claude_desktop_config.json`
   - **Windows:** `%APPDATA%\Claude\claude_desktop_config.json`
2. Add your MCPLambda server:

```json
{
  "mcpServers": {
    "mcpl-server": {
      "url": "https://your-deployment-url.mcplambda.io/mcp"
    }
  }
}
```

---

## 3. Why Streamable HTTP?

MCPLambda has moved away from traditional SSE in favor of **Streamable HTTP** for several reasons:

- **Connection Stability:** Unlike SSE, which requires a single long-lived connection, Streamable HTTP is more resilient to network interruptions and standard proxy timeouts.
- **Firewall Friendly:** It uses standard HTTP patterns that are less likely to be blocked by corporate firewalls or load balancers.
- **Performance:** Reduced overhead for bidirectional communication between your agent and the server.

## Which transport should I choose?

| Client | Recommended Transport | How to Connect |
| :--- | :--- | :--- |
| **Cursor** | **Streamable HTTP** | Use Deployment URL in Settings |
| **VSCode Extensions** | **Streamable HTTP** | Use Deployment URL in Settings |
| **Claude Desktop** | **Streamable HTTP** | Add URL to `claude_desktop_config.json` |
| **Custom Web Apps** | **Streamable HTTP** | Use URL via standard MCP SDKs |

<Cards>
  <Card
    title="CLI Reference"
    href="/docs/cli/"
    description="Manage your deployments from the terminal."
  />
  <Card
    title="Deployment Strategies"
    href="/docs/deployment-strategies/"
    description="Learn more about transport options."
  />
</Cards>