Affordable AI Compute

Access to affordable computational resources is crucial in an AI-driven world. Fair Compute dramatically cuts costs by leveraging a vast network of idle devices, from office laptops to private datacenters, minimizing dependency on expensive cloud services.

Success! Please head out to https://discord.gg/JqjZPwrdZe to stay up to date with latest features and engage with the community.


Start Using it in Seconds

Manage cluster with easy to use API
Serve LLama, Mistral, Stable Diffusion and other models with a single command

## Run ollama server on a community machine
> fair docker run -r nvidia -p 11434:11434 \
    -- ollama/ollama
## Generate an image using stable diffusion
## on a remote GPU-enabled machine
## and download it to your computer
> fair docker run -r nvidia \
    faircompute/stable-diffusion:latest \
    -s "A robot dinosaur" \
    -o /workspace/result.png => result.png
## Use Python API to schedule a task
from fair import FairClient

client = FairClient(user_email="...", user_password="...")
client.run(image='python',
           command=['python', '-c', "print('Hello Fair')"])
## Create a cluster and share it with your team
> fair cluster create my_cluster
> fair cluster -c my_cluster nodes add my_node
> fair cluster -c my_cluster users add my_team@gmail.com

Run Anywhere

Eliminate the need for writing cloud-specific code with Fair API, a unified gateway to major cloud providers, on-prem infrastructure, and customer devices.

Fair Logo
Fair Logo
Fair Logo
Run on-prem
Run in cloud
Run on any device

Share Resources

Empower collaboration with Fair Compute, effortlessly sharing computational capacity with partners. Pool computational resources to enhance utilization, lower costs, and access essential burst capacity when needed.

Shared Computational Resources

Maximize Utilization

Run computations on your computers when not using them with an easy-to-use app. Reduce the money spent on cloud by up to 97%

1

Install the app

Available for Windows, Mac and Linux

2

Choose the time

When you don't use your computer

3

Deploy with Ease

Run the service on your devices for free

3

Scale and save

Leverage community machines at bargain price


Only Pay for What you Need

For many workloads desktop hardware is a great choice. Get a great price from the community for the hardware you need or leverage your own idle computers to get unprecedented savings.

Stable Diffusion

Amazon AWS

NVidia A100
Iterations per second 44.03
1074 iteration per dollar at ≈ $4.1 per hour

Lambda

NVidia A100
Iterations per second 44.03
4003 iteration per dollar at ≈ $1.1 per hour

Fair Community

GeForce RTX 4090
Iterations per second 44.65
11163 iteration per dollar at ≈ $0.4 per hour

Idle Computers

GeForce RTX 4090
Iterations per second 44.65
44652 iteration per dollar at ≈ $0.1 per hour, varies with ⚡ cost