Reserving compute on Ornn means selecting a GPU deployment, specifying your desired start and end dates, choosing how many GPUs you need, and submitting a bid price per GPU-hour. Once submitted, your reservation enters a pending state while the Ornn team reviews and accepts bids. An accepted bid becomes a confirmed reservation that appears in your portfolio.Documentation Index
Fetch the complete documentation index at: https://docs.ornn.com/llms.txt
Use this file to discover all available pages before exploring further.
The full reservation flow
Browse the marketplace
Go to
/marketplace and find a listing that fits your workload. Use the GPU type, operator, quantity, price, and availability filters to narrow down options. Click a row to open the listing detail panel.Select a deployment from /reserve
Navigate to
/reserve to see all available deployments. The page shows a list of GPU deployments grouped into three tabs: All GPUs, Available (open for reservation now), and Upcoming (opening soon). Each card shows the GPU model, quantity, operator, site, region, and an availability or countdown indicator.Use the filter bar to narrow by GPU model, Operator, Location, or Quantity. Click a deployment card to open its reservation page.Open the reservation page
Clicking a deployment card takes you to
/reserve/[deploymentId], the reservation form for that specific deployment. This page shows a calendar view, GPU count selector, bid price input, and a cost summary.Pick your dates on the calendar
The calendar defaults to a week view. Switch to month view for a broader look at availability. Slots shown in the calendar are color-coded:
- Available — the slot is open for reservation
- Booked — the slot is already reserved
- Inactive — outside the deployment’s availability window
anchor_date query parameter to jump to a specific date. The view query parameter accepts week or month.Set your GPU count
Use the GPU count selector to specify how many GPUs you want. You can also set a minimum GPU count (
min_gpu_count) — the lowest number of GPUs you would accept if the full quantity is not available. If min_gpu_count is not set, the full count is required.Enter a bid price
Enter your bid price per GPU-hour in USD. Check the bid summary chart to see the highest bid prices from other tenants over the same period — this helps you gauge a competitive price.
Reservation statuses
After submitting, you can track the status of your reservation in your portfolio at/portfolio/bids.
| Status | Meaning |
|---|---|
| pending | Your bid has been submitted and is awaiting review by the Ornn team |
| accepted | Your bid was accepted; the reservation is confirmed and appears in your portfolio |
| active | The reservation period has started and compute is accessible |
| completed | The reservation term has ended |
Bids are reviewed manually by the Ornn team. Acceptance is not instantaneous. You’ll receive a notification when your bid is accepted and the reservation is confirmed.
Calendar query parameters
The reservation calendar at/reserve/[deploymentId] accepts the following query parameters to control the initial view:
| Parameter | Type | Description |
|---|---|---|
gpu_type | string | Pre-select a GPU type filter |
facility | string | Pre-select a facility filter |
view | week or month | Calendar view mode (default: week) |
anchor_date | date string (YYYY-MM-DD) | The reference date for the calendar display |
What’s next
Understanding bid-based pricing
Learn how bidding works, what fields make up a bid, and how to update or cancel a bid before acceptance.
Completing checkout
After your bid is accepted, complete checkout to confirm your reservation and set up billing.