​

NanoGPT MCP Server

​

Installation

npx @nanogpt/mcp

​

Claude Code

claude mcp add nanogpt --scope user \
  --env NANOGPT_API_KEY=YOUR_API_KEY \
  -- npx -y @nanogpt/mcp

claude mcp add nanogpt --scope user --env NANOGPT_API_KEY=YOUR_API_KEY -- cmd /c "C:\Program Files\nodejs\npx.cmd" -y @nanogpt/mcp

​

IDE Integration

{
  "mcpServers": {
    "nanogpt": {
      "command": "npx",
      "args": ["-y", "@nanogpt/mcp"],
      "env": {
        "NANOGPT_API_KEY": "your_api_key_here"
      }
    }
  }
}

​

Configuration

​

Available Tools

• chat: Send messages to any model supported by NanoGPT.
• image-generation: Generate images using models like Flux, DALL-E 3, etc.
• web-search: Search the web and get back clean, LLM-friendly results.
• scrape-urls: Extract content from one or more websites.
• youtube-transcribe: Get transcripts from YouTube videos.
• balance: Check your current NanoGPT account balance.
• list-models: List available text, image, audio, or video models.
• vision: Analyze images using vision-capable models.

​

Tool Parameters

​

chat

• messages (required): Array of message objects (role + content)
• model (optional): Model ID to use for the request
• temperature (optional): Sampling temperature
• max_tokens (optional): Maximum tokens in the response

​

image-generation

• prompt (required): Text description of the image to generate
• model (optional): Image model ID
• n (optional): Number of images to generate (1–10)
• size (optional): Dimensions (e.g., 1024x1024)
• quality (optional): Quality setting (model-dependent)

​

web-search

• query (required): Search query string
• depth (optional): standard or deep
• fromDate / toDate (optional): Date range filter (YYYY-MM-DD)

​

youtube-transcribe

• urls (required): Array of YouTube URLs (up to 10 per request)

​

Resources

• nanogpt://raw/{id}: Access the raw JSON response from a previous API call.
• nanogpt://image/{id}: Access an image generated during the session.

​

Cost & Billing

• Paid tools deduct from your NanoGPT account balance.
• Model listing operations are complimentary.
• Failed requests (timeouts, errors) are not charged.

​

Troubleshooting

​

API key validation

• Ensure NANOGPT_API_KEY is set in your MCP client configuration.
• Verify the key in your NanoGPT API dashboard and remove any extra spaces.
• If Claude Code fails immediately with Invalid API key, verify the MCP launcher command too, especially on native Windows where npx should run via cmd /c "C:\Program Files\nodejs\npx.cmd".

​

Timeouts for long operations

{
  "mcpServers": {
    "nanogpt": {
      "command": "npx",
      "args": ["-y", "@nanogpt/mcp"],
      "env": {
        "NANOGPT_API_KEY": "your_api_key_here",
        "NANOGPT_TIMEOUT_MS": "900000"
      }
    }
  }
}

​

Node.js version compatibility

node --version

​

MCP client connection checks

• Re-add the server configuration if your client can’t connect.
• Review your MCP client logs for connection errors.

​

Getting Started

1. Get your API key from the NanoGPT API dashboard.
2. Configure your MCP client (like Claude Desktop) with the server settings.
3. Start chatting with your agent and ask it to “Search the web for…” or “Generate an image of…”.