Instances and nodes#
Limit Value Note
GPUs per instance 1 to 8 Depends on the model; the exact sizes are on each model page GPU models 47 NVIDIA and AMD, data-center, workstation and consumer Regions 3 US East, US West, EU Central Instances per account Limited by capacity and balance No hard count; a deployment needs free capacity of that model in that region Instance name 60 characters Optional; a reference like i-7f3a2c is always assigned Multi-node clusters On request Anything wider than one node is arranged as reserved capacity
Volumes#
Limit Value Note
Volume size 10 GB to 4,000 GB In steps of 10 GB, chosen at creation Resize in place Not available Create a larger volume and copy across while both are attached Attachments One instance at a time A volume cannot be shared between two running instances Region Fixed at creation A volume cannot move between regions; copy through an instance Volumes per account No fixed limit Each is billed at $0.08 per GB-month while it exists Deletion Only by you An attached volume must be detached first; nothing is deleted on our initiative
Balance and spending#
Limit Value Note
Minimum top-up $30 No top-up fee; credit does not expire Balance needed to deploy At least one hour of the instance rate The platform will not start what it cannot bill for an hour Billing granularity One started minute Hourly price ÷ 60, no minimum duration Spend cap Optional, monthly Instances are stopped before the cap is crossed; volumes are kept Balance at zero Instances stop Volumes survive and keep being billed; nothing is deleted without contact Payment confirmation On-chain confirmations A payment already sent is credited even if you close the page
Account and security#
Limit Value Note
Failed sign-ins Throttled Repeated failures slow down, then block, attempts from that address Two-factor authentication TOTP, 8 recovery codes Recovery codes are shown once; store them outside the account Sessions Listed per device Revoke one device or all the others from Settings SSH keys Several per account Installed on every instance you launch API keys Created in the console Shown once at creation; revoke and rotate at any time Support tickets Accounts with credit or a live resource Everyone can read the help centre and use a support ticket
API and webhooks#
API and CLI account limits.
Rate limits are set per account when access is enabled. The table below lists the default limits. Contact us about access or limits for your workload.
Limit Value Note
Requests 120 per minute per key Read endpoints; bursts above it answer 429 with Retry-After Write requests 20 per minute per key Creating and stopping instances and volumes Page size 100 items Cursor pagination on every list endpoint Webhook delivery 5-second timeout Retried with a backoff for up to an hour Metadata endpoint No limit Link-local, inside the instance; poll every 5 to 15 seconds
Raising a limit#
Capacity limits are physical, and we will tell you plainly when something is not available. Everything else — a larger volume, more instances at once, a higher API rate, a cluster wider than one node — is a message away. Tell us what you are running, the shape you need and the dates, through a support ticket or a ticket in the console if you already have an account.