GPUs are expensive, so why leave them on? Pay by the second for the predictions you run, based on the hardware the model is run on.
$0.0002 per second
(or, $0.012 per minute)
$0.00055 per second
(or, $0.033 per minute)
16GB GPU RAM
$0.0023 per second
(or, $0.138 per minute)
40GB GPU RAM
Different models run on different hardware. You’ll find the hardware specifications under the "Performance" heading on each model's page on Replicate. For example, see kuprel/min-dalle
If you cancel your prediction before it starts, there’s no charge. If you cancel it after it’s started we stop the prediction immediately and only bill you for the time used so far.
When your prediction completes successfully, we calculate how long it ran for and add it to your account. Once per month we charge you for the time you’ve used. The minimum billable time for any prediction is 1 second. You can see your current usage on your account page.
🚀 Ready to use the API?
To get started, you'll need to sign up and enter your billing info. There is no charge to sign up, and your predictions will be billed by the second.