The RunPod MCP server is an implementation of the Model Context Protocol that connects your AI agent and assistants like Claude, Cursor, etc directly to your RunPod account. It provides structured and secure access to your GPU cloud resources and account settings, so your agent can compare compute options, inspect pods, configure clusters and serverless workloads, and manage templates, credentials, and settings on your behalf.
- Compute selection and pod inspection: Have your agent compare available GPU and CPU types, review pricing and availability, and inspect the details of a specific pod.
- GPU cluster creation: Direct your agent to create GPU clusters for multi-node distributed computing and machine learning workloads.
- Serverless workload configuration: Let the agent create or update serverless workloads with the GPU, scaling, and capacity settings you need.
- Templates and credential management: Instruct your agent to create, update, or remove container templates, manage private container registry access, and store secure secrets.
- Account and security settings: Have your agent retrieve your account information or update user settings such as your SSH public key.