--- title: "Install MCP Tools 2026: 5 Minutes + 6 Servers (Step-by-Step)" description: "From license to first tool call in 5 minutes. Memory + CRM + GEO + Crew + Academy + Suite, for Claude Desktop, ChatGPT and Cursor. With config snippets." author: "Matthias Meyer" published: 2026-02-24 updated: 2026-02-24 language: en tags: ["mcp", "tutorial", "installation", "anleitung", "ai-tools"] canonical: "https://studiomeyer.io/en/blog/mcp-tools-installation-tutorial" markdown_versions: ["https://studiomeyer.io/de/blog/mcp-tools-installation-tutorial.md", "https://studiomeyer.io/en/blog/mcp-tools-installation-tutorial.md", "https://studiomeyer.io/es/blog/mcp-tools-installation-tutorial.md"] publisher: "StudioMeyer, https://studiomeyer.io (llms.txt: https://studiomeyer.io/llms.txt)" --- # Install MCP Tools 2026: 5 Minutes + 6 Servers (Step-by-Step) Installing MCP tools sounds like a tech project. It's not. In five minutes, you'll have your first MCP server configured and ready to use AI tools directly in Claude Desktop, ChatGPT, or any other compatible client. This guide walks you through the entire process -- from understanding to a working connection. ## What Is MCP? MCP stands for Model Context Protocol. It's an open standard introduced by Anthropic in 2024 and since adopted by all major AI providers -- OpenAI, Google, Microsoft, and many more. The idea is simple: instead of every AI tool needing its own integration, there's one unified standard. An MCP server provides tools, an MCP client (such as Claude Desktop) uses them. Communication runs over JSON-RPC 2.0, secured with API keys. For you as a user, this means: **one format, usable everywhere.** An MCP tool configured for Claude works with minimal changes in ChatGPT or other clients too. ## Step 1: Choose a Product Before configuring anything, you need an MCP tool. The [our tools overview](https://studiomeyer.io/en/developers) offers specialized AI tools for different business areas: - **StudioMeyer Memory** -- Multi-agent memory layer across all your AI assistants - **StudioMeyer CRM** -- Customer management directly from your AI chat - **GEO** -- AI visibility tracking, where you are cited in ChatGPT and Claude - **Crew** -- Multi-agent orchestration for specialized workflows - **SmartBot** -- AI chatbot for your website with optional memory bridge Each product comes with its own MCP endpoint and specialized tools. Choose what you need and configure only that. ## Step 2: Get Your License Key After purchase, you'll receive a license key in the format: ``` sm_live_YOUR_KEY ``` This key authenticates all your MCP requests. Keep it secure -- it's your access to the tools. **Important:** The key is used as a Bearer token in your configuration. Don't share it publicly and don't commit it to Git repositories. ## Step 3: Configure Your AI Client ### Option A: Claude Desktop Claude Desktop is Anthropic's native client with built-in MCP support. Configuration is done through a JSON file. **Config path:** - macOS: `~/Library/Application Support/Claude/claude_desktop_config.json` - Windows: `%APPDATA%/Claude/claude_desktop_config.json` **Add the configuration:** ```json { "mcpServers": { "studiomeyer-crm": { "url": "https://crm.studiomeyer.io/mcp", "headers": { "Authorization": "Bearer sm_live_YOUR_KEY" } } } } ``` Replace `sm_live_YOUR_KEY` with your actual license key. The server name (here `studiomeyer-crm`) is freely chosen -- it only serves as identification in your client. After saving: **restart Claude Desktop.** The tools will automatically appear in the tool list. ### Option B: ChatGPT ChatGPT has supported MCP since early 2026. Setup works through Custom GPTs or directly in ChatGPT Plus: 1. Open GPT settings or create a Custom GPT 2. Add a new "Action" 3. Enter the URL: `https://crm.studiomeyer.io/mcp` 4. Select "API Key" as authentication 5. Enter your license key Available tools are automatically discovered and usable in chat. ### Option C: Other Clients Any MCP-compatible client works with the same endpoint. The URL follows this pattern per product: ``` https://{product}.studiomeyer.io/mcp ``` Where `{product}` is `memory`, `crm`, `geo`, or `crew`. Authentication is always a Bearer token with your license key. ## Step 4: Test the Connection After configuration, test the connection. In Claude Desktop, a successful connection means the server's tools appear in the tool list. **Simple test:** Ask your AI client: > "What tools are available from the StudioMeyer MCP server?" The client should list the available tools. For StudioMeyer CRM, that includes tools for contact management, lead tracking, and pipeline management. **If the connection doesn't work:** 1. **Check your key:** Is the API key correct and complete? 2. **Check the URL:** Is the endpoint right? No typos? 3. **Restart the client:** Some clients only load configuration at startup 4. **Firewall:** Make sure outgoing HTTPS connections are allowed ## Multiple MCP Servers Simultaneously You can use multiple MCP servers in parallel. Each gets its own entry in the configuration: ```json { "mcpServers": { "studiomeyer-crm": { "url": "https://crm.studiomeyer.io/mcp", "headers": { "Authorization": "Bearer sm_live_YOUR_CRM_KEY" } }, "studiomeyer-memory": { "url": "https://memory.studiomeyer.io/mcp", "headers": { "Authorization": "Bearer sm_live_YOUR_MEMORY_KEY" } } } } ``` The AI automatically knows which tool belongs to which server and selects the right one for your request. ## Security The connection to the MCP Gateway runs over HTTPS (TLS-encrypted). Your data is not stored or used for training. The API key ensures only you can access your tools. **Best practices:** - Never put license keys in public repositories - Use one key per product - If you suspect compromise: renew the key in your dashboard ## Next Steps Installation is the easy part. Real productivity begins when you integrate the tools into your daily workflow: 1. Start with one tool and learn its available functions 2. Add more tools as you understand the workflow 3. Discover all available products in [our tools overview](https://studiomeyer.io/en/developers) Five minutes of setup. Lasting productivity gains. Getting started with professional MCP tools is that simple. ## Read more, Model Context Protocol - [MCP Marketplaces in April 2026: A Field Report from 33 Platforms](https://studiomeyer.io/en/blog/mcp-marketplaces-2026.md) - [Personal Suite: Email, Calendar, Messaging in One MCP](https://studiomeyer.io/en/blog/personal-suite-mcp-alternative.md) - [What is MCP? The New Standard for AI Tools Explained](https://studiomeyer.io/en/blog/was-ist-mcp-standard.md)