Download and bind a client
Open My Nodes and choose a desktop or CLI client for your system and architecture. Generate a registration token after signing in and follow the client instructions.
Turn spare GPUs into available capacity.
Manage client binding, hardware detection, model deployment and admission in one supplier workspace. Plan capacity for both internal workloads and platform supply.
Detect hardware and deploy
Inspect GPUs, memory and runtime requirements. Select a supported model and variant, deploy it and verify inference.
Complete testing and admission
Review benchmarks, admission status and heartbeats. Use node events and operational records to investigate issues after admission.
Supply and maintenance
Inspect requests and earnings in node details. Drain work before stopping a node for maintenance, and reserve capacity for internal workloads.
SUPPLIER WORKSPACE