Features
Usage Accounting
Understand per-request cost and hourly wallet charges.
Every completion includes a usage object with token counts and cost in USD, plaingateway only.
Every response includes X-PG-Cost-USD, plaingateway only.
The generation lookup returns cost and usage after the request.
Billing posts the difference between each closed hour of measured usage and amounts already charged, every 5 minutes.
Corrections create an adjustment, and an hour whose usage disappears is credited back.
Recorded cases: chat_usage_include and chat_usage_include_stream.