PairX ExchangePrototype · share GPU, get GPU
connecting…
Nodes online–
Shared VRAM–
Models offered–
Jobs done–
Tokens made–
Saved by failover–
Nodes
Try the exchange
The answer appears here, with the node that ran it.
Recent jobs
| Time | Model | Node | Tokens | GPU | Tries | Status |
|---|
Nodes dial out to the exchange over one connection, so contributors open no ports. Jobs go to the smallest free GPU that can run the model; if a node drops mid-job, the job re-runs on another node. Only AI jobs run on nodes, never shell access.