One MCP Server.
130+ AI Models.
Connect Claude Code, Cursor, VS Code, Windsurf, and any MCP-compatible host to RunAPI — generate images, videos, music, and call LLMs without leaving your editor.
{
"mcpServers": {
"runapi": {
"command": "npx",
"args": ["-y", "@runapi.ai/mcp"]
}
}
}What is RunAPI MCP Server?
RunAPI MCP Server connects MCP-compatible coding tools to the RunAPI platform. It lets your AI assistant browse 130+ models across 18 providers, inspect input parameters, check real-time pricing, create media generation tasks, poll task status, check your account balance, and call LLM endpoints — all from within your editor. Discovery tools work without an API key. Authenticated operations require a RunAPI API key.
8 Built-in Tools
Everything your AI assistant needs to discover, price, and run AI models.
list_models
Browse models by modality, service, or action.
get_model_info
Inspect input params, constraints, and pricing for a model.
list_actions
Group endpoint action names by output modality.
check_pricing
Return pricing snapshot for a service + action + model.
create_task
Create a media task and optionally poll until completion.
get_task
Fetch status and latest payload for an existing task.
check_balance
Return account balance and spending metrics.
chat
Send messages to a RunAPI LLM endpoint.
Get Started in 3 Steps
Install
Add the MCP server config to your editor. One JSON block, one command.
claude mcp add runapi -s user -- npx -y @runapi.ai/mcp
Authenticate
Run runapi login to sign in, or create an API key at runapi.ai/dashboard.
runapi login
Use
Ask your AI assistant to generate images, videos, music, or call LLMs. It handles the rest.
"Generate a sunset video with Kling"
130+ Models, 5 Modalities
All models accessible through a single MCP server.
| Modality | Models | Providers |
|---|---|---|
| Image | Flux Kontext, Nano Banana, GPT Image, Seedream, Ideogram, Qwen | Black Forest Labs, Google, OpenAI, Bytedance, Ideogram, Alibaba |
| Video | Kling, Veo, Seedance, Hailuo, Wan, Runway, Luma | Kuaishou, Google, Bytedance, MiniMax, Alibaba, Runway, Luma |
| Music | Suno v4 / v4.5 / v5 / v5.5 | Suno |
| Audio | ElevenLabs TTS, Sound Effects, STT, Isolation | ElevenLabs |
| LLM | Claude, GPT, Gemini, DeepSeek | Anthropic, OpenAI, Google, DeepSeek |
Frequently Asked Questions
Discovery tools like list_models, get_model_info, list_actions, and check_pricing work without an API key. For task creation, balance checks, and LLM chat, you need a free RunAPI API key from the dashboard.
RunAPI MCP Server works with any MCP-compatible host: Claude Code, Codex, Cursor, VS Code (via GitHub Copilot), Windsurf, and Roo Code. It can also be configured with any tool that supports the MCP protocol.
Over 130 models across 18 providers and 5 modalities: image generation (Flux, Nano Banana, GPT Image), video generation (Kling, Veo, Seedance), music creation (Suno), audio (ElevenLabs), and LLMs (Claude, GPT, Gemini, DeepSeek).
Run claude mcp add runapi -s user -- npx -y @runapi.ai/mcp to install globally. For Cursor, Windsurf, and VS Code, use npx @runapi.ai/mcp init <platform> as a fallback. No global npm install required.
The MCP server itself is free and open-source (Apache 2.0). API usage is billed through your RunAPI account with pay-as-you-go pricing. No subscriptions or minimum spend.
Yes. Set the RUNAPI_API_KEY environment variable and configure the MCP server in your agent or CI pipeline. The server runs as a stdio process and does not require a browser or GUI.
Ready to connect your editor to 130+ AI models?
Install the MCP server and start generating in minutes.