If the response is truncated, Amazon Textract returns this token. You can use this token in the subsequent request to retrieve the next set of text detection results.