Connect an MCP client
Every client needs two things: the server URL and your API key in an Authorization header.
URL: https://api.clipbeat.app/mcpHeader: Authorization: Bearer cb_live_...Create the key in Clipbeat Studio, Settings, Developers, API keys (Authentication).
Replace cb_live_... below with your key, or better, keep the key in an environment variable or the client’s secret
store.
Claude Code
Section titled “Claude Code”claude mcp add --transport http clipbeat https://api.clipbeat.app/mcp \ --header "Authorization: Bearer $CLIPBEAT_API_KEY"Check it with claude mcp list, then ask: “Make a preview music video for https://example.com/my-song.mp3”.
Claude Desktop
Section titled “Claude Desktop”Claude Desktop starts local servers from claude_desktop_config.json (Settings, Developer, Edit Config). Use the
mcp-remote bridge to reach the remote server with your key (needs
Node.js):
{ "mcpServers": { "clipbeat": { "command": "npx", "args": ["-y", "mcp-remote", "https://api.clipbeat.app/mcp", "--header", "Authorization:${CLIPBEAT_AUTH}"], "env": { "CLIPBEAT_AUTH": "Bearer cb_live_..." } } }}Restart Claude Desktop after saving. The header is passed through env so the space in Bearer cb_live_... survives on
every operating system.
Cursor
Section titled “Cursor”Add to ~/.cursor/mcp.json (all projects) or .cursor/mcp.json (one project):
{ "mcpServers": { "clipbeat": { "url": "https://api.clipbeat.app/mcp", "headers": { "Authorization": "Bearer cb_live_..." } } }}VS Code
Section titled “VS Code”Add to .vscode/mcp.json. VS Code asks for the key once and stores it securely:
{ "inputs": [ { "type": "promptString", "id": "clipbeat-key", "description": "Clipbeat API key", "password": true } ], "servers": { "clipbeat": { "type": "http", "url": "https://api.clipbeat.app/mcp", "headers": { "Authorization": "Bearer ${input:clipbeat-key}" } } }}The tools appear in Copilot Chat in agent mode.
ChatGPT
Section titled “ChatGPT”ChatGPT connects to MCP servers through a sign-in flow (OAuth), which Clipbeat will support soon. Today, add Clipbeat to a custom GPT as an Action using the REST API:
- In the GPT editor, open Configure, then Create new action.
- Choose Import from URL and enter
https://api.clipbeat.app/v1/openapi.json. - Under Authentication, choose API Key, auth type Bearer, and paste your key.
The GPT can then import songs, start videos and poll their status with the same rules as the MCP tools.
Use the MCP Client Tool node with an AI Agent node:
| Setting | Value |
|---|---|
| Endpoint | https://api.clipbeat.app/mcp |
| Server Transport | HTTP Streamable |
| Authentication | Bearer Auth, with your key as the token |
| Tools to Include | All, or pick the tools you need |
For fixed workflows without an agent, the HTTP Request node calling the REST API is often simpler.
Any Streamable HTTP client
Section titled “Any Streamable HTTP client”The server speaks plain JSON-RPC 2.0 over POST. You can talk to it with nothing more than curl:
curl -s https://api.clipbeat.app/mcp \ -H "Authorization: Bearer $CLIPBEAT_API_KEY" \ -H "Content-Type: application/json" \ -H "Accept: application/json, text/event-stream" \ -d '{"jsonrpc":"2.0","id":1,"method":"initialize","params":{"protocolVersion":"2025-06-18","capabilities":{},"clientInfo":{"name":"my-client","version":"1.0.0"}}}'With the official SDKs:
// npm install @modelcontextprotocol/sdkimport { Client } from '@modelcontextprotocol/sdk/client/index.js';import { StreamableHTTPClientTransport } from '@modelcontextprotocol/sdk/client/streamableHttp.js';
const transport = new StreamableHTTPClientTransport(new URL('https://api.clipbeat.app/mcp'), { requestInit: { headers: { Authorization: `Bearer ${process.env.CLIPBEAT_API_KEY}` } },});const client = new Client({ name: 'my-app', version: '1.0.0' });await client.connect(transport);
const { tools } = await client.listTools();console.log(tools.map((t) => t.name));
const balance = await client.callTool({ name: 'get_balance', arguments: {} });console.log(balance.structuredContent);await client.close();# pip install mcpimport asyncioimport os
from mcp import ClientSessionfrom mcp.client.streamable_http import streamablehttp_client
async def main(): headers = {"Authorization": f"Bearer {os.environ['CLIPBEAT_API_KEY']}"} async with streamablehttp_client("https://api.clipbeat.app/mcp", headers=headers) as (read, write, _): async with ClientSession(read, write) as session: await session.initialize() tools = await session.list_tools() print([t.name for t in tools.tools]) balance = await session.call_tool("get_balance", {}) print(balance.structuredContent)
asyncio.run(main())Troubleshooting
Section titled “Troubleshooting”| Symptom | Cause and fix |
|---|---|
401 on connect |
The key is missing, mistyped or revoked. Check the header is exactly Authorization: Bearer cb_live_.... |
Client tries GET /mcp and gets 405 |
The server does not offer a server-sent event stream. Streamable HTTP clients fall back to POST only; older SSE-only clients are not supported. |
insufficient_credits from a tool |
The account needs credits. Open the topup_url from the result. |
429 |
Too many requests for this key. See Rate limits. |
| Tools do not appear | Restart the client after editing its config. Some clients only load MCP tools in agent mode. |