About OpenGPU
Building the open network for GPU compute.
Compute should not belong to a handful of companies. We connect the world's idle GPUs with everyone who needs them.
Our story
It started with a student and a professor.
A young computer scientist from distributed systems and infrastructure, and a professor focused on distributed algorithms and network theory, kept returning to one question: if the internet can coordinate almost everything globally, why do we still accept bottlenecks and lock-in for compute?
2017
The question
Years before the current AI wave, it was clear that centralized clouds would become a choke point: provisioning delays, rising costs and long GPU queues.
2017onward
From whiteboards to architecture
How do you route AI workloads across independent providers? Verify that work was really done? Keep costs down while making sure work is done? The answer became a decentralized task protocol and a global routing layer.
Build
From two people to a network of builders
Engineers, designers, lawyers, researchers and community builders joined step by step. Investors came later, once it was clear this was long-term infrastructure.
Mar 2025
The network goes live
Real providers, real workloads, from day one.
Today
325+ providers in 40+ locations
Serving inference, training, rendering and research, with Relay bringing the network to enterprises through a standard API.
How we work
Built by people with something to prove.
The same people who design protocols spend full days in support channels, write documentation and explain the architecture on community calls.
Humble work
No task is too small. We fix the edge cases no one sees, because it makes the network stronger.
Long-term focus
The need for open compute isn't going away. We build for the network at a billion in scale.
Shared upside
An ecosystem where builders, providers and early supporters share in the value they create.
People
A global team behind an open network.
Core contributors
Protocol development, security, product, operations, design, growth and support. Many have built together across multiple projects before OpenGPU.
Advisors
Deep expertise in distributed systems, infrastructure and commercial operations, keeping the network grounded in real-world use.
Join us
We're remote-first, small and technical. We value depth over credentials, and applications are reviewed by the team, not recruiters.
Open positions
Contact
Talk to the team.
Questions, feedback or partnership ideas. For enterprise workloads, the pilot program is the fastest route.