- INVENTORY
- CAPACITY
- AUTOMATION
- HUMAN
Short answer
How can I scale model replicas using my AI?
Your AI checks your current model inventory and uses the available tools to adjust replica counts based on your capacity requirements. This keeps your hardware usage aligned with your active model needs.
Scaling outcomes
Where capacity meets demand
What happens when your AI manages your replicas.
INVENTORY
Real-time model counts
The AI pulls your current model list to see exactly what is running. You get an accurate view of your active inventory.
CAPACITY
Balanced hardware use
The AI adjusts replica counts to ensure you aren't over-provisioned or under-powered. Your hardware stays matched to your load.
AUTOMATION
Hands-off scaling
Once you set the logic, the AI handles the math and the execution. You stop manually calculating replica needs.
HUMAN
Manual overrides
The AI handles the routine adjustments, but you retain full control. You decide when to step in and change settings yourself.
The workflow
What your AI does when the request arrives.
The AI handles the heavy lifting of inventory checks and scaling commands.
Audit inventory
The AI looks at your current deployment to see which models are active and how many are running.
nim_list_modelsCalculate needs
Your agent compares the active model list against your capacity targets to find the gap.
nim_list_modelsAdjust replicas
The AI executes the command to increase or decrease the number of replicas for specific models.
nim_scale_replicasVerify state
The AI checks the inventory again to confirm the scaling command finished successfully.
nim_list_models
Try it
Copy these to start.
Use these prompts to begin managing your replicas.
Starting points
These are starting points. Swap out the model names or replica counts for your specific setup.
NVIDIA NIM Connector
You're all set. Choose your MCP client and follow the setup instructions.
https://edge.vinkius.com/vk_preview_70BV0qtRSn6oYi65LbVY4KDoGOvWiulNTIr6cS4G/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 — NVIDIA NIM capabilities are ready to use.
{
"mcpServers": {
"nvidia-nim-mcp": {
"url": "https://edge.vinkius.com/vk_preview_70BV0qtRSn6oYi65LbVY4KDoGOvWiulNTIr6cS4G/mcp"
}
}
}List all my current models and tell me how many replicas are running for each.
Scale the Llama-3-70b model to 5 replicas.
Check my model inventory and tell me if I have enough capacity for a new deployment.
Reduce the replica count for my Mistral model to 2.
Claude
ChatGPT
Cursor
VS Code
Windsurf
Claude Code
JetBrains
Cline
Start here
Connect NVIDIA NIM once, then ask.
Just link your account once. Your credentials stay encrypted, and you can immediately start giving commands to your AI client.
Connect NVIDIA NIM to your AIFAQ
How this task behaves
- 01
Can the AI delete my models?
No. The AI can only scale the number of replicas for existing models.
- 02
Will the AI change my hardware configuration?
No. The AI only adjusts the replica counts of the models themselves, not the underlying hardware settings.
- 03
Does the AI need my login every time?
No. You connect once, and the connection stays active through Vinkius.
- 04
Can I see what models are currently running?
Yes. The AI can pull a full list of your active models and their current replica counts.
- 05
Can the AI scale models I haven't deployed yet?
No. The AI can only scale models that are already present in your active inventory.
More questions about NVIDIA NIM? The Connector page answers them. See everything the NVIDIA NIM Connector can do
Connect NVIDIA NIM to Claude, Cursor, ChatGPT & more
