Choose your guide
GPU clouds
Vast.ai
Instances, templates, and Vast Serverless to Pods and Serverless.
Lambda
On-demand instances, filesystems, and 1-Click Clusters.
CoreWeave
Kubernetes workloads, Slurm, and object storage.
Paperspace and DigitalOcean
Machines, Notebooks, Deployments, and GPU Droplets.
Serverless and inference platforms
Modal
Modal apps, functions, classes, and volumes to the Runpod SDK.
Replicate
Hosted models, Cog models, and deployments.
Baseten
Truss models, Model APIs, and autoscaling settings.
fal
Model APIs, the queue API, and fal Serverless apps.
Together AI
Serverless inference, dedicated endpoints, and fine-tuned models.
Fireworks AI
Serverless inference, on-demand deployments, and custom models.
Hugging Face
Inference Endpoints, custom handlers, and GPU Spaces.
Hyperscalers
AWS
EC2 GPU instances, SageMaker, and Bedrock.
Google Cloud
Compute Engine, Vertex AI, and Cloud Run.
Azure
GPU VMs, Azure Machine Learning, and Azure OpenAI.
Frameworks and API versions
Cog
Run an existing Cog image on Runpod Serverless.
OpenAI
Point OpenAI client code at a vLLM endpoint on Runpod.
Runpod REST v1 and GraphQL
Move existing Runpod integrations to REST API v2.
Flash
Move a
runpod-flash project to the Runpod SDK.Where workloads land
The other platform’s product name matters less than what the workload does. Sort each one by its shape.How a migration goes
Every guide follows the same eight steps, so you can run them in order or jump to the one you need.1
Take inventory
List what runs on the other platform and find the code that calls it. Consoles often hold resources that no file in your repo mentions.
2
Set up Runpod
Create an account, get an API key, and install the tools you’ll use: runpodctl, the Runpod SDK, the MCP server, or Terraform.
3
Choose where each workload lands
Use the table above, then pick a GPU that fits the model. GPU stock varies by data center, so check what’s available before you commit to a type.
4
Move images and code
Most container images run on Runpod unchanged if they’re built for
linux/amd64. Platform-specific contracts, such as a handler signature or a health route, need a small port.5
Move data and models
Copy datasets and checkpoints to a network volume. For models on Hugging Face, cached models skip the copy.
6
Deploy
Create the Pod or endpoint from the console, runpodctl, the REST API, the SDK, or Terraform.
7
Update callers
Change the URL, the auth header, and the request shape in the code that calls the workload. Put the new values behind configuration so you can switch back.
8
Verify, cut over, and shut down
Send real requests to both deployments and compare. Shift traffic, then stop what’s still billing on the other platform.
Migrate with a coding agent
The Runpod plugin for coding agents includes a migration skill,runpod-cloud-migrate. It scans your repository for the platforms it depends on, opens the matching playbook, and proposes a plan before it changes anything. It migrates one workload at a time and verifies each with a real request.
1
Install the Runpod skills and MCP server
- Claude Code
- Other agents
The plugin installs the skills and connects the hosted MCP server:
2
Ask for the migration
Open your project and paste this prompt. In Claude Code, you can run
/runpod:migrate-cloud instead.3
Review the plan
The agent stops after the inventory and plan. Check the targets, the GPU choice, and the list of gaps, then approve it. Nothing billable is created before that point.