The InputTokens property returns the number of input tokens
used by the most recent prompt sent to the model. This value reflects
the number of tokens consumed by the prompt and any associated context
included with the request.
Token counts are determined by the model provider and may vary
depending on the content of the prompt, the model being used and how
the provider processes the request. Token counts do not directly
correspond to characters or words, and the same text may produce
different token counts when used with different models.
If no prompt has been sent, or if the request failed before the
provider returned token usage information, this property will return
zero.