REQUENTA / FAQ
A closer look
at the market.
The market model, from a supplier’s available window to a buyer’s delivered compute.
The market
Who is Requenta for?+
AI teams with repeat training, fine-tuning, batch evaluation and inference workloads. On the supply side: professional operators and holders of GPU reservations they can resell, with firm capacity windows.
Why can spot compute cost less?+
Prices ease when available capacity exceeds demand. Flexible workloads can choose those windows. As demand tightens, a higher quote can attract more supply by clearing the next seller’s net floor.
What is the product?+
GPU-hours attached to a qualified resource group and a firm delivery window. Requenta brings booking, isolated execution and billing into one interface across providers.
What does the index describe?+
The index design groups comparable delivered transactions by hardware, topology, region and service class. It gives context to market prices. Your accepted quote determines the price of your booking.
Buying compute
Can my price change during a run?+
The booking model fixes your rate for the hours you accept. Changes in the market affect new quotes. Extra hours require a new quote and an available window.
Which hardware does Requenta focus on?+
The initial market design centres on dedicated 8×H100 SXM nodes in Europe. A compatible offer specifies GPU configuration, network topology, region, service class and a firm delivery window.
How do I discuss a pilot?+
Share your GPU requirements, runtime, region and expected volume through the workload form. We’ll follow up to discuss capacity fit and commercial terms. Submitting an enquiry does not make a booking.
What is the billing model?+
Meter the compute actually delivered, reconcile usage and consolidate it on one invoice. Requenta’s fee model is 5.5% of completed third-party transaction value.
Supply & execution
Can I offer capacity I have reserved?+
Yes, if you can resell it and commit the full delivery window. Supplier qualification covers hardware, performance and isolation, as well as your right to offer that capacity.
How is supplier infrastructure protected?+
The execution design uses scoped identities, isolated workspaces, network and storage boundaries, expiring access and verified cleanup. The first integration path is an adapter for authorised capacity on an existing Kubernetes cluster.
Who controls availability and price?+
The supplier defines the available pool, minimum net price and firm hand-back time. A booking must clear that floor after all deductions. Internal workloads outside the authorised pool stay under the supplier’s control.
How do supplier incentives work?+
The incentive design allocates a share of higher-priced trades to active suppliers according to monthly delivered hours in the same market. This rewards the regular supply that supports repeat bookings. Only completed, non-refunded external trades count.
Recurring workloads
What are Credits?+
A prepaid GPU-hour model for recurring workloads. Discuss your expected volume, budget and usage pattern to assess the commercial fit.
What is Anchor?+
A monthly-volume model for teams that want to discuss a compute budget, price ceiling and flexibility requirements. Workload fit and availability form part of the agreement.