- MONITOR
- CONTROL
- AUDIT
- HUMAN
Short answer
How can I manage my Vast.ai instances with AI?
Your AI client uses the Vast.ai API to pull a list of your active rentals and kill specific ones on command. This lets you monitor usage and stop billing through a chat interface. You get full visibility and control over your running GPUs.
Fleet management outcomes
What you can do with your fleet
The AI handles the API calls so you can focus on your workloads.
MONITOR
Real-time visibility
The AI pulls your current instance list so you know exactly what is running. You get a clear snapshot of your active compute.
CONTROL
Instant termination
When you're done with a job, tell your AI to kill the instance. It stops the billing immediately by removing the rental.
AUDIT
Usage tracking
Your AI can review your active instances to help you spot runaway processes. This prevents unexpected costs from forgotten rentals.
HUMAN
Manual oversight
You remain the final decision maker. The AI identifies the instances, but you decide which ones to kill to avoid losing progress.
The workflow
What your AI does when you request management.
The AI acts as a bridge between your chat interface and the Vast.ai API.
Scan active rentals
The AI checks your account to see every GPU instance currently active in your fleet.
list_instancesIdentify targets
It parses the list to find the specific ID or name of the instance you want to address.
list_instancesExecute shutdown
Once you confirm, the AI sends the command to remove the instance from the cloud.
delete_instanceVerify removal
The AI re-scans your list to ensure the instance is gone and billing has stopped.
list_instances
Try it
Copy these to start.
Use these prompts to begin managing your GPUs.
Starting points
These are basic commands. Swap out the instance IDs or names with your actual Vast.ai data.
Vast.ai Connector
You're all set. Choose your MCP client and follow the setup instructions.
https://edge.vinkius.com/vk_preview_FDhKvj0GzXpDt4zTlC7GXQudBJQAnkTGnbdwdCBo/mcpClaude Desktop
Follow the steps below to connect in seconds.
- 1In Claude Desktop, open Settings → Connectors.
- 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
- 3Click Add and start a new chat — Vast.ai capabilities are ready to use.
{
"mcpServers": {
"vastai-gpu-rental-cloud-api-mcp": {
"url": "https://edge.vinkius.com/vk_preview_FDhKvj0GzXpDt4zTlC7GXQudBJQAnkTGnbdwdCBo/mcp"
}
}
}List all my active GPU instances on Vast.ai.
Which of my instances are currently running?
Kill the instance with ID 123456.
Find my running instance named 'training-node-01' and terminate it.
Claude
ChatGPT
Cursor
VS Code
Windsurf
Claude Code
JetBrains
Cline
Start here
Connect Vast.ai once, then ask.
Just link your account via the Vinkius dashboard. Your credentials stay encrypted and your AI client gets immediate access to your fleet.
Connect Vast.ai to your AIFAQ
How this task behaves
- 01
Can the AI create new GPU instances?
No. This MCP only supports listing and deleting existing instances.
- 02
Will the AI delete my instances without asking?
No. The AI identifies the instances, but you must explicitly tell it to delete one.
- 03
Can I see how much money I've spent?
The AI can list your instances, but it does not have a tool to access your billing history or total account balance.
- 04
Does this work with any AI client?
Yes, it works with any MCP-compatible client like Claude, Cursor, or Windsurf.
- 05
How do I stop being billed?
You must tell the AI to delete the specific instance you are using to stop the billing cycle.
More questions about Vast.ai? See everything the Vast.ai Connector can do
Connect Vast.ai to Claude, Cursor, ChatGPT & more
