Inference cost

[TheMrguiller]

Hi,

I would like to know if there is any feature available to track the exact cost of analyzing documents passed to LangExtract. Since the provider’s response (including token usage) is handled internally during processing, it seems that this information is not exposed or easily traceable.

Is there any way to collect or retrieve the cost (e.g., token usage or API cost) for each document processed?

Thanks in advance!