Private GPUs and model choices.

How it works

Managed for you

Use Gee without looking after computers. We run the shared hardware for you.

Private GPU

Choose hardware reserved for your business. This is a separate service from the shared hardware every plan runs on. The add-on is $999 a month, plus Multi at $199 a month per location. GPU time costs extra, at our cost plus 10%. Your quote sets the hardware and hourly rate before you agree. Other services keep their own usage charges.

Run your own GPUs

Connect a GPU you own or supply through your customer-operated compatible endpoint. You manage the hardware and use models under their licence terms.

Where Gee runs

Choose managed computing, a private GPU dedicated to your business, or a GPU you own or supply.

Hardware

Private dedicated hardware is rented, not owned by you. Run your own GPUs through a customer-operated compatible endpoint, with models you are licensed to use.

How it is billed

Private GPU requires Multi at $199 a month per location, plus a $999 monthly add-on for setup, your private model and management. GPU hours cost extra, with no token meter. Your quote sizes the GPUs for your locations and sets the hourly rate. The plan and other usage are billed separately. A single small location usually does not need a private GPU. GPU hours are billed on top at actual cost plus 10% per GPU-hour, with no token meter.

For businesses that need dedicated computing. It is not a promise of a lower bill or a better answer.