AI Prosumer
EN

ShareAI AI Gateway

One request.
Your choice of compute.

Connect your app once. Run compatible AI requests on your own machine, our decentralized network or connected AI inference providers.

Follow a request

Your app asks.
ShareAI connects.

Scroll to follow the request into ShareAI. Then choose where it runs.

Your app’s requestOne API integration
ShareAIChoose where it runs
One integration. More than one way to run.

Choose a route above to see the request travel. You can try all three.

Interactive routing illustration. No AI request is sent.

Mix and match

Use AI.
Provide it, too.

Your own infrastructure and the wider network can work together. Use the route that fits each job, and put eligible spare capacity to work as an AI Prosumer.

Explore AI Prosumer

For Community and Company Providers.

01

When you need your machine.

Run compatible requests on infrastructure you control.

02

When you need more capacity.

Choose network capacity or a connected provider. You control which routes and fallback your setup allows.

03

When your infrastructure is idle.

Share eligible compute to earn through paid contributions, or exchange it for AI usage under the program you choose.

Model support, availability and contribution terms depend on your setup and program.

The published catalog

See our live
catalog rates.

Compare published input and output prices. Open a model for its provider options and pricing details.

Explore all models and rates
Published starting prices in EUR per one million tokens.
ModelInput fromOutput from
openai/gpt-6-astraShareAI€9.1€45.5
openai/gpt-5.4ShareAI€2.21€13.3
openai/gpt-4o€2.55€10

Starting prices can come from different offers. Open a model to check its provider, price tiers and access terms.

From idea to inference

A clear path
for every request.

Your application owns the experience. ShareAI connects its model requests to the compute you choose.

  1. 01

    Choose your model.

    Check what it can do, where it is available and how it is priced.

  2. 02

    Connect your app.

    Follow the API guide for authentication and your first inference request.

  3. 03

    Choose your route.

    Set provider preferences and decide whether another route may be used.

  4. 04

    Keep track.

    Review usage and costs in the ShareAI account that funds the requests.

Start with the API guide

For model creators

Bring your models to the Gateway.

You choose how your models participate, including who can serve them. Model ownership and serving infrastructure remain separate choices.

Explore the Creator Program

A few useful answers

Before your
first request.

Can I use more than one route?

Yes. Use supported routes for different requests and workloads. Model compatibility, your selected providers and your configured fallback rules determine where a request can run.

Does using my own machine keep everything local?

Your machine can run the supported model call. ShareAI still routes that call, and your app may use its own storage, tools or services. If model execution must stay on your infrastructure, configure the route to prevent external fallback.

Can I share capacity as an individual or a company?

Explore Community Provider participation for personal hardware and Company Provider participation for business infrastructure. Supported hardware, model rights and the contribution program determine what you can share and how it is rewarded.

Explore the Provider Program
Is the routing illustration sending a real request?

It is an interactive preview. It does not connect a machine, change a provider setting, spend credits or send a prompt.

Where can I use a ready-made app?

Explore apps in the ShareAI Marketplace. The Gateway is the integration layer for applications; each app provides its own experience and features.

Explore Marketplace apps

ShareAI AI Gateway

Your app.
Your next possibility.