0%
Loading content
N
Private
Run open
models,
privately
Nyxos routes each prompt to an independent GPU node running an open model. No account, no tracking, nothing logged but the ledger entry.
Decentralized
No gatekeepers,
no
lock-in
Inference runs on independent nodes, not one company's servers. Pick any open model. Payment settles peer to peer, per request.
Network
A grid
of GPU
nodes
Every node is real hardware running an open model. A prompt routes to a free node, gets answered, and the node is paid per job. That is the network.
Open
The models
you actually
choose
Llama, Qwen, Gemma, DeepSeek and Phi. Open weights, served on hardware you do not have to buy, behind one OpenAI-compatible API.
Earn
Put
your GPU
to work
Run one command to join as a provider. Your GPU answers prompts and earns credits per job, redeemable for $NYXOS at launch.
Free
Private AI.
Run free.
Open models, peer to peer settlement, no accounts and no logging. AI that belongs to the people who run it.