Open AI, powered by people.
Chat with open models like Llama and Qwen, running on GPUs that people around the world share with the network. No big data center in the middle.
How it works
Eranos X connects people with spare graphics cards to people who want to use AI.
People share their GPUs
Hosts run our app. When their computer is idle, it answers questions with an open model through Ollama.
You ask a question
We send it to a free, verified GPU with the right model and stream the answer straight back to you.
We keep checking
Every GPU is approved by our team and quietly tested with questions we know the answers to. Hosts that fail are removed.
Trust and privacy, honestly
A shared network needs straight answers about who can see what.
- Every GPU is approved and testedHosts are reviewed by hand, then must pass hidden test questions before they answer anyone, and are re-tested at random.
- We don't keep your conversationsOur servers pass messages through without storing them. Your chat history stays in your own browser.
- Only approved open modelsHosts can only serve models on our list, so you always know what you're talking to.
The GPU answering you can see your message.
Your question has to reach a person's computer to be answered. Hosts agree not to look at or keep anything, and we test them, but we can't make it impossible. So please don't share passwords, health or money details, or anything private.
Have a GPU that sits idle?
Share it with people who can't afford AI subscriptions. Mac with Apple Silicon or a PC with an NVIDIA card.