Aerostack
Aerostack
electrical_services

Runpod MCP Server — Hosted Api Connectors Integration

MCP Server language Hosted language Public

GPU cloud compute via RunPod — deploy pods, manage GPU instances, start and stop workloads for AI training and inference.

aerostack @aerostack verified
• v0.1.0 • MIT • Updated Jun 29, 2026
add_circleAdd to Workspace
robot_2

Use with AI AssistantsMCP

Connect Claude, Cursor, or any MCP-compatible client — then call tools directly

① Add This MCP Server

Paste into your AI client config — then all its tools are available instantly.

.claude/mcp.json
{
  "mcpServers": {
    "runpod": {
      "url": "https://mcp.aerostack.dev/s/aerostack/mcp-runpod",
      "headers": {
        "Authorization": "Bearer YOUR_AEROSTACK_TOKEN"
      }
    }
  }
}

Replace YOUR_AEROSTACK_TOKEN with your API token from the dashboard.

② Call a Tool

Ask your AI assistant to call a specific tool, or send raw JSON-RPC:

+2 more

Natural Language Prompt

“Use the _ping tool to verify runpod credentials by calling a lightweight read endpoint. used internally by aerostack to validate credentials”

Using a Workspace?

Add this MCP to your Workspace — your team shares one token, secrets are stored securely, and every AI agent in the workspace can call it without per-user setup.

add_circleAdd to Workspace

description Overview

mcp-runpod — Runpod MCP Server

GPU cloud compute via RunPod — deploy pods, manage GPU instances, start and stop workloads for AI training and inference.

Live endpoint: https://mcp.aerostack.dev/s/aerostack/mcp-runpod

What You Can Do

This MCP server gives AI agents access to Runpod via 7 tools. Connect it to any Aerostack workspace and your agents can interact with Runpod directly.

Available Tools

Tool Description
list_gpu_types List all available GPU types on RunPod with memory, pricing (spot and on-demand), and availability in secure vs community cloud.
list_pods List all pods in your RunPod account with their status, image, and machine details.
get_pod Get detailed information about a specific RunPod pod including status, GPU count, and logs setting.
create_pod Deploy a new GPU pod on RunPod. Creates a secure cloud pod with the specified GPU type and container image.
stop_pod Stop a running RunPod pod (pauses billing while preserving the pod configuration and volume).
resume_pod Resume a stopped RunPod pod to restart billing and execution.
terminate_pod Permanently terminate and delete a RunPod pod. All data not on persistent volume will be lost.

Configuration

Variable Required Description
RUNPOD_API_KEY Yes Your RUNPOD API KEY from the service's developer settings

Quick Start

Add to Aerostack Workspace
  1. Go to aerostack.dev → Your Project → MCPs
  2. Search for "Runpod" and click Add to Workspace

Add the following secrets under Project → Secrets:

  • RUNPOD_API_KEY

Once added, every AI agent in your workspace can use Runpod tools automatically.

Direct API Call
curl -X POST https://mcp.aerostack.dev/s/aerostack/mcp-runpod \
  -H 'Content-Type: application/json' \
  -H 'X-Mcp-Secret-RUNPOD-API-KEY: your-runpod-api-key' \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"list_gpu_types","arguments":{}}}'

License

MIT

terminal Tools (8)

Available tools on this MCP server. Each tool can be called directly from any AI agent.

terminal
_ping #1

Verify RunPod credentials by calling a lightweight read endpoint. Used internally by Aerostack to validate credentials.

terminal
list_gpu_types #2

List all available GPU types on RunPod with memory, pricing (spot and on-demand), and availability in secure vs community cloud.

terminal
list_pods #3

List all pods in your RunPod account with their status, image, and machine details.

terminal
get_pod #4

Get detailed information about a specific RunPod pod including status, GPU count, and logs setting.

terminal
create_pod #5

Deploy a new GPU pod on RunPod. Creates a secure cloud pod with the specified GPU type and container image.

terminal
stop_pod #6

Stop a running RunPod pod (pauses billing while preserving the pod configuration and volume).

terminal
resume_pod #7

Resume a stopped RunPod pod to restart billing and execution.

terminal
terminate_pod #8

Permanently terminate and delete a RunPod pod. All data not on persistent volume will be lost.

Details

upgrade Version 0.1.0
gavel License MIT
wifi Transport streamable-http
lock Access Public
category Category API Connectors
terminal Tools 8

language Live Endpoint

https://mcp.aerostack.dev/s/aerostack/mcp-runpod

Sub-50ms globally · Zero cold start

Publisher

aerostack
@aerostack verified

Pre-built functions for the most common MCP tool patterns. Clone, extend, and deploy.

Tags

runpod gpu cloud ai-training inference
Browse more servers

More in API Connectors

Browse API Connectors MCPs →
hub

Aerostack Registry

by @aerostack

Discover and invoke any MCP, Function, or Skill published to the Aerostack marketplace — the universal AI capability hub.

api-connectors
star 0
hub

Algolia

by @aerostack

Search indexes, manage records, browse data, and configure ranking in Algolia — AI-native instant search access.

api-connectors
star 0
hub

Arangodb

by @aerostack

Query documents, run AQL, traverse graphs, and manage collections in your ArangoDB database — AI-native multi-model database access.

api-connectors
star 0
hub

Ayrshare

by @aerostack

Post, schedule, and analyze social media across 13 platforms — Facebook, Instagram, X, LinkedIn, TikTok, Bluesky, Threads, Reddit, Pinterest, YouTube, Telegram, Snapchat, Google Business.

api-connectors
star 0
hub

Basecamp

by @aerostack

Manage projects, to-dos, messages, schedules, and campfire chats in Basecamp — AI-native project management.

api-connectors
star 0
hub

Bigquery

by @aerostack

Run SQL queries, list datasets and tables, inspect schemas, and export results from Google BigQuery — AI-native data warehouse access.

api-connectors
star 0

Frequently asked questions

What is the Runpod MCP server and what can it do? +

The Runpod MCP server is hosted on Aerostack and exposes these tools to your AI agent: `_ping`, `list_gpu_types`, `list_pods`, `get_pod`, `create_pod`. You get one hosted URL — no self-hosting — that works from Claude, Cursor, ChatGPT, Gemini, VS Code, or any MCP-compatible client, and you can share it with your team or combine it with other MCP servers in a workspace.

Is the Runpod MCP server hosted, or do I have to run it myself? +

It's hosted on Aerostack's edge infrastructure — you don't deploy or maintain anything. Add it to a workspace and you get one authenticated URL, with secrets encrypted, that any AI agent or editor can connect to. Use it solo or share the same URL across your whole team.

Which AI agents and editors can use the Runpod MCP server? +

Any MCP client: Claude and Claude Code, Cursor, ChatGPT, Gemini, Windsurf, Cline, VS Code, and custom agents. Because it's one hosted URL, the same Runpod MCP server works everywhere — and you can compose it with other MCP servers, skills, and functions behind a single workspace URL.

How do I install the Runpod MCP server in Claude Desktop? +

Add the following to your Claude Desktop config (`claude_desktop_config.json`): ```json { "mcpServers": { "@aerostack/mcp-runpod": { "command": "npx", "args": ["-y", "@aerostack/@aerostack/mcp-runpod"] } } } ``` Then restart Claude Desktop and the tools will appear automatically.

How do I use the Runpod MCP server in Cursor? +

In Cursor, open **Settings → MCP** and add: ```json { "name": "@aerostack/mcp-runpod", "command": "npx", "args": ["-y", "@aerostack/@aerostack/mcp-runpod"] } ``` Save and reload Cursor. The MCP tools will be available in Agent mode.

Does Runpod MCP require authentication? +

Yes. Runpod requires authentication. Check the MCP's documentation for the required credentials.