mcp · wavespeedai — Claude Code MCP server: generation and setup
Connect Claude Code to WaveSpeed’s model catalog and ask an agent to generate images, video, audio, or 3D content. The server can also estimate costs and track running jobs.
claude mcp add --transport http mcp https://api.wavespeed.ai/mcpWhat mcp does
The server searches WaveSpeed’s current catalog, returns a model’s input schema, and submits generation requests. You can check an estimated price before running a model. Other tools show account balance, past predictions, and spending; billing details require a key with Billing access.
Long-running jobs return a prediction ID, which the agent can use with get_prediction to check progress. For a local file, use an input value such as @./photo.jpg so the server uploads it. A bare path passes through unchanged. The hosted server cannot access files on your computer, so provide a public URL or upload through a client that can send files.
For Claude Code, the remote setup takes one command. Local use needs a WaveSpeed key in WAVESPEED_API_KEY or a saved session from wavespeed login; without authentication, model runs cannot start.
What it helps with
- search the WaveSpeed catalog for a model
- generate image, video, audio, or 3D content
- estimate a model run’s cost before submitting it
- check a prediction’s status and outputs by ID
Example prompts for the agent
- Create a hero image with a city at night.
- Replace the background in
@./photo.jpgwith a bright kitchen. - Add subtle parallax motion to
@./hero.pngand estimate the price first.
How to install mcp
1Platform CLI: remote server
claude mcp add --transport http mcp https://api.wavespeed.ai/mcp2Platform CLI
claude mcp add --env WAVESPEED_API_KEY=<WAVESPEED_API_KEY> --env WAVESPEED_BASE_URL=<WAVESPEED_BASE_URL> --env WAVESPEED_CLIENT_NAME=<WAVESPEED_CLIENT_NAME> --transport stdio wavespeedai-mcp -- npx -y @wavespeed/mcpReplace <…> with your own values
1Platform CLI: remote server
codex mcp add wavespeedai-mcp --url https://api.wavespeed.ai/mcp2Platform CLI
codex mcp add wavespeedai-mcp --env WAVESPEED_API_KEY=<WAVESPEED_API_KEY> --env WAVESPEED_BASE_URL=<WAVESPEED_BASE_URL> --env WAVESPEED_CLIENT_NAME=<WAVESPEED_CLIENT_NAME> -- npx -y @wavespeed/mcpReplace <…> with your own values
Config file
{
"mcpServers": {
"wavespeedai-mcp": {
"url": "https://api.wavespeed.ai/mcp"
}
}
}{
"mcpServers": {
"wavespeedai-mcp": {
"command": "npx",
"args": [
"-y",
"@wavespeed/mcp"
],
"env": {
"WAVESPEED_API_KEY": "<WAVESPEED_API_KEY>",
"WAVESPEED_BASE_URL": "<WAVESPEED_BASE_URL>",
"WAVESPEED_CLIENT_NAME": "<WAVESPEED_CLIENT_NAME>"
}
}
}
}Replace <…> with your own values
Config file
{
"servers": {
"wavespeedai-mcp": {
"type": "http",
"url": "https://api.wavespeed.ai/mcp"
}
}
}{
"servers": {
"wavespeedai-mcp": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@wavespeed/mcp"
],
"env": {
"WAVESPEED_API_KEY": "<WAVESPEED_API_KEY>",
"WAVESPEED_BASE_URL": "<WAVESPEED_BASE_URL>",
"WAVESPEED_CLIENT_NAME": "<WAVESPEED_CLIENT_NAME>"
}
}
}
}Replace <…> with your own values
1Platform CLI: remote server
gemini mcp add -t http wavespeedai-mcp https://api.wavespeed.ai/mcp2Platform CLI
gemini mcp add -e WAVESPEED_API_KEY=<WAVESPEED_API_KEY> -e WAVESPEED_BASE_URL=<WAVESPEED_BASE_URL> -e WAVESPEED_CLIENT_NAME=<WAVESPEED_CLIENT_NAME> wavespeedai-mcp npx -y @wavespeed/mcpReplace <…> with your own values
Step by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about mcp
How do I install mcp in Claude Code?
To connect the hosted server, run claude mcp add --transport http mcp https://api.wavespeed.ai/mcp.
What does mcp do?
It connects an agent to WaveSpeed models so it can inspect model inputs, submit generations, and retrieve results. It also provides price estimates and account usage tools.
Can the hosted server read files on my computer?
No. Give it a public URL or upload the file through a client that supports file transfer.
Similar MCP servers
All MCP servers for AI and ML →mcp by other authors comes first.
