# Overview

The DeGirum *AI Hub* is a cloud platform for quickly prototyping edge AI applications. It provides cloud access to the *Device Farm*, a collection of real SoCs and AI accelerators hosted by DeGirum. This allows you to run model inference without the time and cost of hardware bring-up, either directly in the browser with the *Inference Dashboard* or programmatically through DeGirum *PySDK*.

*Public Model Zoos* offer curated, pre-compiled models for multiple supported hardware targets, so you can run them immediately without packaging or setup.

*Workspaces* add advanced capabilities such as role-based collaboration, the ability to bring and manage your own models, and device management.

The *Cloud Compiler* produces artifacts for supported hardware targets, which are saved in *Workspace Model Zoos*. AI Hub also includes a *Fleet Manager* for device oversight and administration pages for managing *Workspace Tokens*, *Workspace Settings*, and *Tasks*.

AI Hub integrates with PySDK, so applications prototyped in the cloud can be deployed on local hardware with the same code, typically requiring only changes to the device or host settings and artifact name. 

<a href="https://hub.degirum.com/?utm_source=docs.degirum.com&#x26;utm_medium=site&#x26;utm_campaign=ai-hub-overview" class="button primary">Sign up for AI Hub</a> <a href="quickstart" class="button secondary">Get started</a>

## AI Hub Features

<table data-view="cards"><thead><tr><th></th><th></th><th data-hidden data-card-cover data-type="files"></th><th data-hidden data-card-target data-type="content-ref"></th></tr></thead><tbody><tr><td><strong>Hardware Explorer</strong></td><td>Begin prototyping without hardware setup or maintenance</td><td><a href="https://content.gitbook.com/content/iNUrjcsNMTjUhiNPIF4u/blobs/OMcCKLzbIbLwqQh5hPCL/Device%20Farm.png">ai-hub--device-farm.png</a></td><td><a href="hardware-explorer">hardware-explorer</a></td></tr><tr><td><strong>Inference Console</strong></td><td>Run inference with your browser, no setup required</td><td><a href="https://content.gitbook.com/content/iNUrjcsNMTjUhiNPIF4u/blobs/r8BvnCfvmEt823CatSUG/Browser%20Inference.png">ai-hub--browser-inference.png</a></td><td><a href="hardware-explorer">hardware-explorer</a></td></tr><tr><td><strong>Public Models</strong></td><td>Begin developing with pre-configured models</td><td><a href="https://content.gitbook.com/content/iNUrjcsNMTjUhiNPIF4u/blobs/RaTchEuPMNiNImbI2kh9/Model%20Zoo.png">ai-hub--model-zoo.png</a></td><td><a href="public-models">public-models</a></td></tr><tr><td><strong>Cloud Compiler</strong></td><td>Port custom models to different hardware platforms</td><td><a href="https://content.gitbook.com/content/iNUrjcsNMTjUhiNPIF4u/blobs/XBnNtaWoHHJXQKSz8ERx/Model%20Compiler.png">ai-hub--model-compiler.png</a></td><td><a href="workspaces/cloud-compiler">cloud-compiler</a></td></tr><tr><td><strong>PySDK Integration</strong></td><td>Use PySDK to run inference on scalable edge deployments</td><td><a href="https://content.gitbook.com/content/iNUrjcsNMTjUhiNPIF4u/blobs/2q4omXs0UZ6WChXuX6Jq/PySDK%20Integration.png">ai-hub--pysdk-integration.png</a></td><td><a href="pysdk-integration">pysdk-integration</a></td></tr></tbody></table>