Trust & transparency
Trust & transparency
A low price means little if you receive a different model. Unbar gives you the model that you request, with public rates and usage records that you can examine.
Our commitments
The model you request is the model we serve
Unbar serves the model you select in your API request. We do not silently replace it with a cheaper model. A different model is your choice.
Model information and prices are public
Our catalog shows model names, identifiers, makers, and detailed input, output, and available cache rates. Compare prices before you request access.
Usage records you can compare with your bill
After you get access, you can review recorded usage and charges in the console. Compare them with the published rates for each billed token type, including applicable cache charges.
Zero prompt retention
Unbar does not store your API prompts or completions and does not use them to train models. Operational records, such as token counts and charges, are separate from conversation content.