Skip to main content
Use token preflight to estimate prompt size before a paid call.
The response returns total_tokens and the selected routed model.