> ## Documentation Index
> Fetch the complete documentation index at: https://docs.comfy.org/llms.txt
> Use this file to discover all available pages before exploring further.

# Deploying ComfyUI

> There are many ways to deploy ComfyUI: Serverless API, Comfy Cloud, and Self Hosting

<CardGroup cols={3}>
  <Card title="Serverless API" icon="cloud-arrow-up" href="/development/serverless/overview">
    Deploy a versioned ComfyUI environment as a managed endpoint with on-demand GPU capacity.
  </Card>

  <Card title="Comfy Cloud" icon="cloud" href="/development/deploy/cloud">
    A managed, always-available ComfyUI. No environment to build, models are pre-installed.
  </Card>

  <Card title="Self-Host" icon="server" href="/development/deploy/self-hosting">
    Run ComfyUI on your own hardware, a rented GPU, or a container. Full control.
  </Card>
</CardGroup>

## Which deployment should I use?

|                           | Serverless API                                                                   | Comfy Cloud                                       | Self-Host                                                          |
| ------------------------- | -------------------------------------------------------------------------------- | ------------------------------------------------- | ------------------------------------------------------------------ |
| **What it is**            | Your ComfyUI build (models, custom nodes), deployed as a serverless API endpoint | Fully managed ComfyUI with limited customizations | ComfyUI wherever you run it: your machine, Runpod, Vast.ai, Docker |
| **GPUs**                  | Managed, on-demand, scale within bounds you set                                  | Managed for you                                   | You provide and manage them                                        |
| **Models & custom nodes** | Whatever you put in the Build                                                    | Pre-installed and managed by Comfy                | You install and manage them                                        |
| **Scaling**               | Configure how you want to scale (0 - 1000s of GPUs)                              | Autoscaling                                       | Manual                                                             |
| **Endpoint**              | `https://<deployment>.run.comfy.app`                                             | `https://cloud.comfy.org`                         | Your server URL, e.g. `http://localhost:8188`                      |
| **Cost model**            | Pay for GPU capacity while deployed                                              | Subscription (API access requires a paid tier)    | Your hardware or provider bill                                     |
| **Best for**              | Production apps that run a pinned workflow at scale                              | Getting started fast, no infrastructure to run    | Full control over hardware, models, and data                       |

## Choosing in practice

* **You want full control over your Deployment**: Use the [Serverless API](/development/serverless/overview). You define what custom nodes, models, python dependencies you want to use. The infrastructure is all managed for you, but still configurable.
* **You want to skip infrastructure entirely.** Use [Comfy Cloud](/development/deploy/cloud). It is effectively a deployment that is always there: models are pre-installed and the endpoint never goes away.
* **You already have GPUs, or need hardware and models under your control.** [Self-host](/development/deploy/self-hosting), on your own machine or on a GPU cloud such as Runpod or Vast.ai.

## Next step: run workflows against it

Once a deployment is live, see [Running Workflows](/development/run-workflows/overview) to call it from your application. The [Comfy SDKs](/development/api-development/sdks) work against all three targets; self-hosted instances additionally need the [API Proxy](/development/comfyui-server/api-proxy) during the v2 beta.