runapi-mcp — Claude Code MCP server: what it does, how to install
Connects coding assistants to the RunAPI catalog for generating images, video, music, and speech. Discovers models and checks pricing directly in chat.
claude mcp add --transport http runapi-mcp https://mcp.runapi.ai/mcpWhat runapi-mcp does
This server bridges MCP-compatible developer tools with the RunAPI media platform. It gives your coding assistant access to a catalog covering text-to-speech, image synthesis, video rendering, and music creation. The assistant inspects schema parameters, checks live runtime pricing, and searches prompt templates before submitting any task.
Browsing the embedded model list needs no authentication. Generating media or fetching account balances requires credentials via RUNAPI_API_KEY or the login tool. The server acts as a thin client and never runs heavy AI workloads locally.
You can run it via stdio using npx or point directly to the remote hosted endpoint at https://mcp.runapi.ai/mcp. The hosted route skips local Node.js dependencies entirely.
What it helps with
- inspect input schemas and runtime pricing across media generation models
- search reusable prompt templates for creative tasks
- trigger image, video, voice, or music generation from the editor
- track task execution status and review account spending limits
Example prompts for the agent
- List available RunAPI image generation models along with their parameter constraints.
- Find prompt examples for generating an app icon and inspect the recommended model.
- Check runtime pricing for video models and generate a five-second clip.
How to install runapi-mcp
1Platform CLI: remote server
claude mcp add --transport http runapi-mcp https://mcp.runapi.ai/mcp2Platform CLI
claude mcp add --env RUNAPI_API_KEY=<RUNAPI_API_KEY> --transport stdio runapi-mcp -- npx -y @runapi.ai/mcpReplace <…> with your own values
1Platform CLI: remote server
codex mcp add runapi-mcp --url https://mcp.runapi.ai/mcp2Platform CLI
codex mcp add runapi-mcp --env RUNAPI_API_KEY=<RUNAPI_API_KEY> -- npx -y @runapi.ai/mcpReplace <…> with your own values
Config file
{
"mcpServers": {
"runapi-mcp": {
"url": "https://mcp.runapi.ai/mcp"
}
}
}{
"mcpServers": {
"runapi-mcp": {
"command": "npx",
"args": [
"-y",
"@runapi.ai/mcp"
],
"env": {
"RUNAPI_API_KEY": "<RUNAPI_API_KEY>"
}
}
}
}Replace <…> with your own values
Config file
{
"servers": {
"runapi-mcp": {
"type": "http",
"url": "https://mcp.runapi.ai/mcp"
}
}
}{
"servers": {
"runapi-mcp": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@runapi.ai/mcp"
],
"env": {
"RUNAPI_API_KEY": "<RUNAPI_API_KEY>"
}
}
}
}Replace <…> with your own values
1Platform CLI: remote server
gemini mcp add -t http runapi-mcp https://mcp.runapi.ai/mcp2Platform CLI
gemini mcp add -e RUNAPI_API_KEY=<RUNAPI_API_KEY> runapi-mcp npx -y @runapi.ai/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 runapi-mcp
How to install runapi-mcp in Claude Code?
For hosted access without local runtime dependencies, run claude mcp add --transport http runapi-mcp https://mcp.runapi.ai/mcp. To run locally via stdio with an API key, use claude mcp add --env RUNAPI_API_KEY=<RUNAPI_API_KEY> --transport stdio runapi-mcp -- npx -y @runapi.ai/mcp.
What does runapi-mcp do?
It exposes tools for searching AI models across images, video, speech, and music, checking live pricing, and running generation tasks through RunAPI.
Do I need an API key to browse models?
Catalog discovery, parameter checks, and prompt searches work without credentials. An API key or sign-in is required only when you generate media or check your account balance.
