Partner with us
Hosting

Your GPUs, earning while you sleep

List idle hardware on the OpenLink marketplace. We bring the customers, run the billing, and answer the support tickets — you keep control of pricing and capacity.

20,000+GPUs on platform
50+Providers on platform
24/7Customer support
YouSet your own prices
What you get

Everything around the hardware

Running GPUs is a hardware problem. Selling GPU hours is a demand, billing, support and compliance problem. We take the second half.

Demand and support

Enterprise customers already searching the marketplace for capacity. We handle discovery, onboarding, billing and 24/7 support — you never speak to a customer unless you want to.

Your terms, your prices

Set your own rate. Offer on-demand, interruptible or reserved. Configure GPU, CPU, storage and bandwidth billing separately. You're the operator; we're the platform.

Utilisation you can see

Live occupancy, revenue per card and payout history, per machine and per rack. Know which SKUs are earning before the invoice tells you.

Who hosts with us

From one workstation to a full hall

01 — Data centers & colo

Verified capacity

Run flagship GPUs with verified status, priority placement in search, and direct introductions to enterprise buyers looking for reserved blocks.

Apply for verification

02 — GPU farm operators

Turn a fleet into a business

You already have the racks and the power contract. Presell upcoming capacity to our sales team so the cards have paying customers the week they land.

Start hosting

03 — First-time hosts

Start with what you have

A single workstation is a legitimate starting point. List it, see what it earns, and decide whether the next rack is worth financing.

How it works

Division of labour

What we handle, what you handle

You focus on the hardware. We focus on everything that turns it into revenue.

OpenLink handles

  • Customer acquisition and marketplace exposure
  • 24/7 customer support and dispute handling
  • Billing, invoicing, collections and payouts
  • Job scheduling and placement across the fleet
  • Marketing and demand generation
  • Fraud screening and customer verification

You handle

  • Hardware uptime and maintenance
  • Network connectivity and bandwidth
  • Power, cooling and physical security
  • Initial hardware setup and the host client

The support line is the part hosts tell us they value most — it's also the part they most underestimate before they start.

Sell the capacity before the hardware lands

Planning a deployment? Share the GPU mix and the delivery timeline with our sales team. We'll match you against enterprise customers looking for reserved blocks — so the cards have paying tenants from the day they rack.

How it works

Three steps from hardware to revenue

Step 01

List your hardware

Install the host client. Your machines appear on the marketplace within minutes, with their real specs and benchmarks.

Step 02

Set your terms

Price the cards, choose which rental types you accept, and configure how storage and bandwidth are billed.

Step 03

We bring the customers

Your GPUs surface in marketplace search and in the API. Revenue accrues automatically; payouts run on a fixed schedule.

FAQ

Common questions

How do I start hosting?
Apply through the form below, install the host client on each machine, and set your prices. Machines appear in marketplace search once they've passed a reliability and bandwidth check.
What GPUs can I list?
Consumer and datacenter cards both have demand — RTX 4090 and 5090 move constantly, and H200 and B200 attract reserved bookings. Older Ampere silicon still earns on price-sensitive workloads. If you're unsure whether a specific SKU is worth listing, ask us before you buy.
How is pricing set?
You set it. The marketplace shows customers what comparable machines are charging, so pricing well above the market means idle cards and well below means you're leaving money on the table. Most hosts start near the median and adjust on observed occupancy.
How do payouts work?
Revenue accrues per second of rented time and is visible in your dashboard as it happens. Payouts run on a fixed schedule to the account on file. Exact terms — schedule, minimum threshold, platform fee — are in your host agreement.
What are the requirements?
Stable power, adequate cooling, and a network connection with enough upstream bandwidth to move datasets and checkpoints. Reliability is measured continuously, and it affects where your machines rank in search — the customer-facing incentive is uptime.
Do I have to deal with customers?
No. Support tickets, billing disputes and abuse reports come to us. You're contacted when something is genuinely a hardware problem — a card throttling, a node dropping off the network.
Can I get hardware financed?
We work with financing partners who can factor projected platform earnings into the assessment. This is arranged case by case rather than as a standard product — talk to the team with your intended GPU mix and timeline.
Apply

Tell us what you're running

GPU mix, node count, location and when it's available. We'll come back with what it's likely to earn on the marketplace and what onboarding involves.

Ready to start earning?

Idle cards cost the same as running ones. List what you already have and find out what it's worth.