GetTokensIn

Returns the number of input tokens consumed by a conversation — the size of the question, the system prompt and any attached files or recordset, as measured by the model. Useful for cost monitoring. Returns 0 for an unknown handle or before an answer has been received.