---
title: "Developer Guide"
description: "What the ARX platform does for model providers and for ISVs, and where each of them starts"
canonical_url: "https://developer.ardge.com/en/guide/overview"
---
# Developer Guide

> What the ARX platform does for model providers and for ISVs, and where each of them starts

**You Build the AI. We Handle Deployment.**

When GPU servers move on-premises, deployment and operations become the challenge. ARDGE unifies GPU infrastructure, inference runtimes, service delivery, and remote management—helping teams move from validation to production faster.

The result: on-premises AI that is easier to deploy, govern, and scale.

## Platform value

**From Model to Service, on One Platform.**

ARX OS combines no-code deployment, Model Hub, App Market, multiple inference runtimes, API key management, OpenAI-compatible endpoints, and remote operations. Whether you are deploying a model, RAG system, AI agent, or complete application, ARDGE provides a clearer path from POC to production.

<card-group>
<card icon="i-lucide-layers" title="Integrated Private AI Platform">

Hardware, operating system, inference runtimes, model services, security, and connectivity—pre-integrated.

</card>

<card icon="i-lucide-repeat" title="Built for Repeatable Delivery">

Standardized deployment, access control, backup, and remote management across different GPU systems and customer environments.

</card>

<card icon="i-lucide-gauge" title="Faster Path to Production">

Reduce model and application deployment from hours to as little as 15 minutes.

</card>
</card-group>

This guide is written for two audiences. Pick the one that matches what you ship.

## For model providers

ARX is the platform your models reach users through — you deploy and configure them from the interface, with no per-site setup work.

- **Deploy from the interface** — publish the model to your Hugging Face repository, then find it in AI Model Center and deploy it in a few clicks
- **Ship deployment settings with the model** — pick the inference engine and its parameters up front, so users deploy the configuration you validated instead of tuning it themselves
- **Serve applications through one gateway** — a deployed model answers on the ARX model gateway in both OpenAI and Claude API format, so any application on the system can call it

## For ISVs

ARX manages the hardware and the system underneath. You ship an image and a template; the platform installs your application on the customer's site, keeps it updated, and connects it to the models, network, and storage it needs.

- **Package an image and it runs** — wrap your application as a container image and describe it in an ARX app template; ARX takes care of the machine, the GPU, and the system underneath
- **Users install with a click** — your application appears in App Market alongside the rest, and users install it from there
- **Version management and remote updates** — publish a new version and users update in place, which keeps post-sales maintenance and day-to-day operations manageable
- **Integrations you would otherwise build yourself** — declare the models your application recommends, its network settings, and the external storage it mounts, and let ARX wire them up

## Where to start

<card-group>
<card icon="i-lucide-brain-circuit" title="Model providers" to="https://developer.ardge.com/en/guide/model-provider/overview">

Deploy your models on ARX and make them available to every application on the system

</card>

<card icon="i-lucide-package" title="ISVs" to="https://developer.ardge.com/en/guide/isv/overview">

Package your application, ship it through App Market, and use the platform's model, network, and storage integrations

</card>
</card-group>


## Sitemap

See the full [sitemap](https://developer.ardge.com/sitemap.md) for all pages.