Configure Product
A product determines which providers, models a customer can use and which usage and rate limit applies to them.
The Free product
When you create a project, Preternatural also creates a Free product.
The Free product:
- Has no SKU.
- Appears as an included product in the dashboard.
- Cannot be deleted.
Add a Product
To add another product:
- Open the project in the Preternatural dashboard
- Go to Products
- Select ‘Add product’
- Enter a display name
- Enter a SKU
- Select a billing interval
- Select Add
Configure a Product
From the Products table, click on the expand icon on a product’s row. This opens the product’s configuration dialog, where you connect provider accounts and set usage and rate limits
Connecting provider accounts
Under Provider accounts section, consists of list of already configured provider accounts that can be connected with the product. More info on configuring provider accounts
Configuring Usage limits
Under the Usage Limit section
Usage Limits can be added by clicking on ‘Add Usage Limit’ button
You can link Usage Limit record with models by either regex pattern matching or selecting the model using it’s model identifier.
Usage Limit can be set based on :
- Requests
- Tokens
- Characters
- Audio File size
- Audio duration
Configuring Rate Limits
Under the Rate Limit section
Rate Limits can be added by clicking on ‘Add Rate Limit’ button
You can link Rate Limit record with models by either regex pattern matching or selecting the model using it’s model identifier.
Rate Limit can be set based on :
- Requests
- Tokens
- Characters
- Audio File size
- Audio duration
Rate Limit duration window can be:
- Per minute
- Per hour
- Per day
Product Customer relation
- A given customer can have access to multiple Preternatural products. Limiting the customer to have access to only one product at a time needs to handled at the application level.
- Customer by default has access to the Free Preternatural Product.
- Assigning a product to customer can only be done by using
client-sdk.
- *Characters, Audio File size, Audio duration metrics are only supported for TTS models