Pricing calculation

AiPrice uses different libraries to be able to calculate the token count in a prompt. Our API will then apply the pricing per token per the model to the token count. The pricing can be calculate as such:

pricing = token_count * models_price_per_token[model_name]

Take for example the following prompt:

This train went from east to west in under 30 seconds. Where did the train go?

This prompt contains 18 tokens. If we were to use the gpt-3.5-turbo model, it will cost us $0.000002 per token. The calculation will be:

pricing = 18 * 0.000002 # equals $0.000036

Keep in mind that the same model used in a different way may cost differently. For example, using the Babbage model for training may cost you $0.0000006 per token while using it for usage will cost $0.0000024 per token.

All these calculations are included in our API.