GPU Cloud Platform

Deploy GPU Servers with one command

Spin up AI/ML infrastructure — Ollama, vLLM, ComfyUI, n8n — directly from Claude Code or API. Powered by NVIDIA RTX 5090.

See How It Works
claude — mcp
claude > Deploy Ollama with llama3.2 on a GPU server
Connecting to JasCloud MCP...
Creating invoice: Ollama GPU Server x 1x RTX 5090
Payment confirmed. Starting provisioning...
Ollama endpoint: http://185.x.x.x:11434 — model: llama3.2
NVIDIA RTX 5090
OpenStack
MCP Protocol
Instant Deploy

Two ways to deploy

Whether you're building AI agents or automating workflows — we've got the infrastructure ready.

MCP Server

Deploy GPU-accelerated AI services directly from Claude Code. Describe what you need in natural language — we handle the rest.

  • Ollama — run any open LLM locally
  • vLLM — high-throughput inference API
  • ComfyUI — image generation workflows
  • Jupyter — GPU notebooks for research
  • Custom Heat templates supported

n8n Ready Instances

Pre-configured workflow automation servers. Connect 400+ integrations with GPU-powered AI nodes out of the box.

  • n8n pre-installed and configured
  • GPU access for AI workflow nodes
  • Persistent storage included
  • Auto-backups & monitoring
  • One-click deploy, zero setup

Everything your business needs in one platform

Deploy GPU-accelerated AI workflows with n8n. LLMs, image generation, speech recognition — no ML experience required.

AI Chatbots & Agents

Build customer support bots, sales assistants, and internal AI agents. Connect to Slack, WhatsApp, or your own app.

Ollama vLLM n8n

Image & Video Generation

Auto-generate product photos, marketing visuals, and thumbnails. Connect to Google Sheets, CMS, or any trigger.

ComfyUI Stable Diffusion

Speech & Transcription

Real-time calls, meetings, podcasts. Generate subtitles, summaries, and action items — all on GPU.

Whisper TTS

Simple, transparent pricing

Pay only for what you use. No hidden fees. Scale up or down anytime.

VM Small
Development & testing
$15/mo
or $0.02/hour
  • 2 vCPU cores
  • 4 GB RAM
  • 40 GB NVMe SSD
  • No GPU
  • n8n or custom stack
4x RTX 5090
Large models & training
$499/mo
or $0.69/hour
  • 32 vCPU cores
  • 128 GB RAM
  • 500 GB NVMe SSD
  • 4x NVIDIA RTX 5090
  • vLLM / training workloads

From zero to GPU in 60 seconds

Three steps. No YAML. No SSH. Just tell Claude what you need.

1

Connect MCP

Add JasCloud MCP server to your Claude Code config. One line, done.

2

Describe Your Need

"Deploy Ollama with Mistral 7B" — Claude picks the right GPU, flavor, and config.

3

Complete Payment

Finish the secure billing step and JasCloud will start provisioning your workload automatically.

Dashboard

Active Services
Credit Balance
Unpaid Invoices
Service
Status
Price
Hostname
Loading...

n8n

Workflow automation platform. Connect APIs, automate tasks, build AI agents — no code required.

Port: 5678

Ollama

Run LLMs locally on GPU. Supports Llama 3, Mistral, Gemma, Phi and 100+ models out of the box.

Port: 11434

Jupyter

Interactive notebooks for ML/AI research. Pre-configured with PyTorch, CUDA and popular libraries.

Port: 8888

ComfyUI

Node-based Stable Diffusion GUI. Build complex image generation workflows with drag-and-drop.

Port: 8188

vLLM

High-throughput LLM inference engine. OpenAI-compatible API with PagedAttention for production serving.

Port: 8000
Invoice #
Status
Amount
Due Date
Loading...

API Key for MCP

Use an API key instead of email/password. Add it to your .mcp.json and Claude Code will auto-authenticate — no login tool needed.

Loading...
Loading...
Ticket #
Subject
Status
Last Updated
Loading...

1 Add to Claude Code

Add this to your MCP server config:

.mcp.json
{
  "mcpServers": {
    "jascloud": {
      "type": "http",
      "url": "/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_API_KEY"
      }
    }
  }
}

2 Use in Claude Code

Just tell Claude what you need in natural language:

Log in to JasCloud with my email
Deploy an Ollama server with Mistral 7B
Show my active services and invoices
Order a VM Small for n8n automation

Available Tools

jas-login

Authenticate with email + password

jas-account

View account details, balance and services

jas-products

List available GPU products and pricing

jas-order

Order a new VM or GPU server

jas-services

List your services and their status

jas-invoices

View invoices and payment status

jas-pay

Pay an invoice with account credit

jas-balance

Check credit balance and unpaid invoices

jas-add-funds

Get a link to add funds to your account

jas-help

Show all available commands

Endpoint

MCP Server
/mcp