RunPod

RunPod

RunPod is one registered infrastructure adapter. Elastic deployments use SkyPilot to provision RunPod Pods; Serverless deployments use RunPod’s native endpoint lifecycle. InferCrane’s base production stack does not require RunPod—enable it explicitly with compose.production.runpod.yaml.

Elastic
Serverless

infercrane doctor --cloud
infercrane plan Qwen/Qwen3-8B --cloud runpod --gpu L40S
infercrane deploy Qwen/Qwen3-8B \
  --cloud runpod \
  --gpu L40S \
  --min 1 \
  --max 4

infercrane doctor --serverless
infercrane deploy Qwen/Qwen3-8B \
  --compute serverless \
  --cloud runpod \
  --gpu L40S \
  --max 4

Both paths can create billable resources. Inspect existing pods and endpoints before retrying a provider operation, retain the original idempotency key, and verify provider inventory after deletion.

See provider setup for credentials and the serverless lifecycle for worker-zero behavior.

GET STARTED

BUILD

OPERATE

INTELLIGENCE

INTEGRATE

REFERENCE

PROJECT

RunPod

​RunPod

RunPod