# Does WaveSpeed have an MCP server?

**Yes — Official.** WaveSpeed maintains an official MCP server.

## Facts

- Verdict: Official (Yes)
- Access: Local server — Runs on your machine over stdio, launched by your MCP client.
- Package: [`@wavespeed/mcp`](https://www.npmjs.com/package/@wavespeed/mcp) on npm
- Repository: [WaveSpeedAI/mcp-server](https://github.com/WaveSpeedAI/mcp-server) — ★ 33, last push 2026-09-29, MIT
- Maintainer: WaveSpeedAI
- Language: TypeScript
- Category: [Art & Culture](https://mcpyet.com/category/art-culture/)
- Verified against vendor docs: 2026-09-29
- GitHub data refreshed: 2026-09-29 (re-checked every 6 hours)

## What it does

> WaveSpeed MCP server — run any model on the live catalog (image, video, audio, 3D) with schema introspection, @path file upload, and price quotes
> — how [WaveSpeedAI/mcp-server](https://github.com/WaveSpeedAI/mcp-server) describes itself

Project site: <https://wavespeed.ai>

## The receipts

- **Official package:** [`@wavespeed/mcp`](https://www.npmjs.com/package/@wavespeed/mcp) on npm. The package page carries the exact launch arguments and required credentials.
- **Official repository:** [WaveSpeedAI/mcp-server](https://github.com/WaveSpeedAI/mcp-server) — ★ 33, pushed 2026-09-29, MIT (active)

## How to connect it

Runs on your machine over stdio, launched by your MCP client.

### Claude Code

```sh
claude mcp add wavespeed -- npx -y @wavespeed/mcp
```

Most servers need an API key passed as an environment variable — check the README for the exact variable names. Add `--scope user` to make it available in every project.

### Claude Desktop

Open **Settings → Developer → Edit Config** and add the server to `claude_desktop_config.json`, then restart Claude Desktop:

```json
{
  "mcpServers": {
    "wavespeed": {
      "command": "npx",
      "args": [
        "-y",
        "@wavespeed/mcp"
      ],
      "env": {
        "API_KEY": "…"
      }
    }
  }
}
```

### Cursor

Add to `.cursor/mcp.json` in your project, or `~/.cursor/mcp.json` for every project:

```json
{
  "mcpServers": {
    "wavespeed": {
      "command": "npx",
      "args": [
        "-y",
        "@wavespeed/mcp"
      ],
      "env": {
        "API_KEY": "…"
      }
    }
  }
}
```

### VS Code

Add to `.vscode/mcp.json`, then pick the server from Copilot's agent-mode tools menu:

```json
{
  "servers": {
    "wavespeed": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "@wavespeed/mcp"
      ]
    }
  }
}
```

### ChatGPT

ChatGPT connects to **remote servers only**, and this one runs on your machine. Either expose it through an MCP gateway that gives it a public URL, or use a client with native local support (Claude or Cursor).

## How we know

Published to the MCP Registry under WaveSpeed's verified namespace ai.wavespeed, pointing to the WaveSpeedAI/mcp-server repo in WaveSpeed's GitHub org and the npm package @wavespeed/mcp (run with npx -y @wavespeed/mcp and a WaveSpeed API key); there is no hosted endpoint.

- GitHub data refreshed 2026-09-29, then every 6 hours.
- Vendor docs verified on 2026-09-29.
- Corrections: hello@mcpyet.com.

## Also in Art & Culture

- [fal.ai](https://mcpyet.com/mcp/fal-ai.md) — Official (Yes)
- [Kling](https://mcpyet.com/mcp/kling.md) — Community only (Sort of)
- [Kie.ai](https://mcpyet.com/mcp/kie-ai.md) — Community only (Sort of)
- [Blender](https://mcpyet.com/mcp/blender.md) — Community only (Sort of)
- [Canva](https://mcpyet.com/mcp/canva.md) — Official (Yes)
- [Higgsfield](https://mcpyet.com/mcp/higgsfield.md) — Official (Yes)

---

Source: https://mcpyet.com/mcp/wavespeed/ — data refreshed 2026-09-29