Language models are used to detect information in documents and make predictions for information extraction.

Third-party large language model (LLM) providers are available for Now Assist skills and AI agents in addition to Now LLM Service. For more information on LLMs in Now Assist, see Manage AI models.

For each Now Assist in Document Intelligence use case, only one LLM can be enabled at a time. Now Assist employs the selected LLM when processing documents for the use case.

For image files that need optical character recognition (OCR) to detect the text in them, OCR models are used to support different language groups. The language selected during use case setup helps the OCR model to detect text in the images.