Get customers for GPUs you aren't using

We bring the customers and place your capacity at the best price, reserved or on demand: the whole cluster, or only the hours you leave idle.

The whole cluster goes on the live index, where customers reserve it for a term or take it on demand. One listing, one price, and we close each deal with you.

Contact information

Company nameWork emailPhone number (optional)

Cluster 1

GPU info

GPU typeNumber of GPUsPrice GPU/hr ($)

Cluster interface

Networking & infrastructure

Interconnect networkNode RAM (GB)Cores per nodeNVMe storage (GB)Geographic locationCloud service provider

Scheduling & availability

Start dateEnd dateMinimum bookable GPUsMin bookable duration (weeks)

Additional details

Two ways to list capacity

You reserved it.
List the hours you don't run.

Keep the hours you actually run
Customers take the idle hours on demand
Priced against the live index, shown below

You run the racks.
Publish what's free.

Whole clusters or bookable slices, for a term or on demand
Your company name stays off the public listing
One form, one desk, no portal to learn

Today's going rate

Live medians across active listings: what customers compare your ask against.

B300: $4.43/GPU-hr; 3,542 offered
B200: $3.70/GPU-hr; 2,368 offered
H200: $3.10/GPU-hr; 1,120 offered
H100: $2.35/GPU-hr; 416 offered