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

# Overview

> A new way to monetize your apps and device fleets.

Turn unused computing capacity into a revenue stream. ZeroGPU enables app developers and device operators to earn by contributing available compute to a global AI inference network.

Whether you reach thousands of users through an app or manage a fleet of powerful gaming PCs, ZeroGPU connects that capacity to AI workloads and handles model delivery, task routing, and orchestration.

Our mission is to orchestrate the world's idle compute to accelerate AI. As an Edge Provider, you help build that network and earn from the inference your devices serve.

## For app developers: add compute monetization

Your app's device footprint can support a new source of revenue alongside advertising, subscriptions, and in-app purchases.

Integrate the ZeroGPU Edge SDK into your application. With users' consent, compatible devices contribute available computing capacity and run supported AI inference tasks locally.

ZeroGPU handles distributing models and assigning work. You manage your integration and track activity and earnings through the [Edge Portal](https://edge.zerogpu.ai).

Supported integration options include [Android apps](https://docs.zerogpu.ai/edge/sdk/android.md), [websites](https://docs.zerogpu.ai/edge/sdk/browser.md), [Chrome extensions](https://docs.zerogpu.ai/edge/sdk/chrome-extension.md), [Telegram Mini Apps](https://docs.zerogpu.ai/edge/sdk/telegram.md), and [supported desktop applications](https://docs.zerogpu.ai/edge/sdk/electron.md).

## For fleet operators: put existing hardware to work

Gaming PCs, university computer labs, esports facilities, and other managed device fleets often have capacity available between sessions or outside their busiest hours.

Connect compatible machines using the [ZeroGPU desktop agent](https://docs.zerogpu.ai/edge/agent.md) and make that capacity available for AI inference.

For example, an esports school can contribute available compute from its gaming PCs when they are not needed for classes or practice. A university can connect eligible lab machines under its device-use policies.

Start with a few devices, measure utilization and earnings, and expand across your fleet.

## How it works

**1. Create your provider account**

Register in the [Edge Portal](https://edge.zerogpu.ai) and choose your contribution method.

**2. Connect your devices**

Integrate an SDK into your app or install the agent on machines you own or are authorized to manage.

**3. Contribute available compute**

Eligible devices receive supported models and run inference tasks assigned by ZeroGPU.

**4. Track activity and earnings**

Monitor connected devices, completed work, and revenue through the portal.

## Global devices. AI closer to demand.

Your devices add both computing capacity and geographic reach to the network. ZeroGPU considers location, compatibility, availability, and load when assigning supported workloads, helping bring inference closer to where it is needed.

## How you earn

You earn from eligible inference work served by your connected devices under your provider terms. Earnings depend on demand, device capabilities, availability, and completed work. Connecting a device makes it available for tasks; it does not guarantee continuous utilization.

## Start monetizing your compute

Visit the Edge Portal to register your app, download an agent, and connect your first devices.

> [card] [Edge Portal](https://edge.zerogpu.ai)
>
> Register your app or fleet.

> [card] [Download the agent](https://apps.zerogpu.ai/edge)
>
> For Windows, macOS, and Linux.

> [card] [Quickstart](https://docs.zerogpu.ai/edge/quickstart.md)
>
> Connect your first devices.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.