Your addressnot connected — showing demo
Your hardware
Detecting GPU… run the dcloud-plugin on your machine to onboard (NVIDIA/CUDA or Apple/Metal supported).
Verification · two-way trust
100.0%
probe pass-rate (recent)
sample-rate p
100%
clean streak
0
0
availability
0.0%
0.0%
Services · share this machine across workloads
100% allocatedfully allocated
dCloud · 100%ready
dHost
0%Host a model, serve inference
µMDL/1K tok
dCloud
100%Rent raw GPU into the pool
µMDL/epoch
dShard
0%Serve a slice of a big model
µMDL/1K tok
dTrain
0%Contribute to community training
µMDL/round
dStorage
0%Store verified vector memory
µMDL/GB·day
Split by % of a GPU (MPS/MIG slicing) or by whole GPUs on a multi-GPU rig. You're paid for verified availability per service — the network probes each slice and pays per verified epoch/job.
Rent out this GPU · Serve marketplace
Rate (µMDL / hr)
Region
GPUs
Payout goes to your connected address — listing is gated by your pairing code. Renters run any workload (docker/miner/model); you're paid per verified epoch. Appears on the Serve page.
Running now
Idle — waiting for the network to route work.
Active lease (dCloud rental)
No active lease.
Probe history
Earnings · payouts