gpu.nz MCP

An official MCP server provided by gpu.nz that allows AI agents to interact with the gpu.nz cloud GPU rental platform. It enables agents to search for available GPUs, estimate VRAM requirements, rent servers, manage instances (start, stop, destroy), view logs, and handle account billing. This allows autonomous agents to provision and manage cloud GPU infrastructure for AI workloads directly.

Endpoint and access

No sign-in: your client connects and lists tools immediately, with no account or key.

No sign-inEndpoint checked 2026-10-26Operated by gpu.nz

MCP endpoint (Streamable HTTP)

https://gpu.nz/mcp

Vendor documentation

What you can do with it

Search for available GPU models and filter by price, memory, and reliability
Estimate VRAM requirements for specific AI models like Llama or Qwen
Rent a server with a pre-configured template such as ComfyUI or vLLM
Stop or destroy running instances to manage costs and free up resources

Tools advertised by the last live check:

list_gpu_models
search_offers
get_offer
list_templates
estimate_vram
get_account
list_instances
get_instance
create_instance
stop_instance
start_instance
destroy_instance
get_instance_logs
list_ssh_keys
add_ssh_key
topup_link

Connect from your client

Each snippet registers the same endpoint; pick the client you use.

This server is operated by the named third party. Listing and endpoint checks are not an endorsement or a security audit. Review the operator's permissions, terms, and data handling before connecting. How listings are checked and corrected.

Claude Code

Run this once in a terminal. It registers a Streamable HTTP server. Use --scope project to write .mcp.json for the repo.

Claude Code configuration

claude mcp add --transport http gpu-nz-mcp https://gpu.nz/mcp

Codex

Run this once. Codex stores MCP servers in ~/.codex/config.toml (or .codex/config.toml in a trusted project).

Codex configuration

codex mcp add gpu-nz-mcp --url https://gpu.nz/mcp

Cursor

Merge this server into .cursor/mcp.json in your project, then enable gpu-nz-mcp in your MCP settings.

Cursor configuration

{
  "mcpServers": {
    "gpu-nz-mcp": {
      "url": "https://gpu.nz/mcp"
    }
  }
}

Cline

Merge this into cline_mcp_settings.json (Cline → MCP Servers → Configure). The type must be streamableHttp for a remote HTTP server.

Cline configuration

{
  "mcpServers": {
    "gpu-nz-mcp": {
      "type": "streamableHttp",
      "url": "https://gpu.nz/mcp",
      "disabled": false
    }
  }
}

VS Code

Merge this server into .vscode/mcp.json. Use MCP: List Servers to start gpu-nz-mcp and inspect its tools.

VS Code configuration

{
  "servers": {
    "gpu-nz-mcp": {
      "type": "http",
      "url": "https://gpu.nz/mcp"
    }
  }
}

Gemini CLI

Run this once. It writes .gemini/settings.json in the current project; pass --scope user for ~/.gemini/settings.json.

Gemini CLI configuration

gemini mcp add --transport http gpu-nz-mcp https://gpu.nz/mcp

Claude Desktop

Merge this into claude_desktop_config.json (macOS: ~/Library/Application Support/Claude/, Windows: %APPDATA%\Claude\), then restart Claude Desktop.

Claude Desktop configuration

{
  "mcpServers": {
    "gpu-nz-mcp": {
      "url": "https://gpu.nz/mcp"
    }
  }
}

Windsurf

Merge this into ~/.codeium/windsurf/mcp_config.json, then refresh MCP in Cascade. Windsurf remote servers use serverUrl, not url.

Windsurf configuration

{
  "mcpServers": {
    "gpu-nz-mcp": {
      "serverUrl": "https://gpu.nz/mcp"
    }
  }
}

Ground your agent in cited data

gpu.nz MCP gives your agent tools. Koskamo gives it reviewed, cited facts to check its answers against. Add it as a second MCP server; public reads need no API key.

Connect Koskamo Browse datasets

See all 217

Do you run gpu.nz MCP?

Prove ownership with a DNS record, a well-known file, or your Official MCP Registry entry, and your edits publish immediately. Proofs are re-checked weekly; admins can still hide or remove a listing.

Claim this listing