Supported Models for Generative AI in SuiteScript

The following table lists the models supported when using the llm.generateText(options) and llm.generateTextStreamed(options) functions, along with their associated capabilities. Starting in NetSuite 2026.2, additional models and capabilities have been introduced to enhance functionality.

Supported Models and Capabilities

Supported Capabilities Overview

• Retrieval-augmented generation (RAG): This feature allows you to provide documents with text content using the options.documents parameter. Such documents enhance the LLM's responses by providing additional context. The LLM will return llm.Citation objects, signifying the source of the information it used. More information about RAG can be found in the Oracle Cloud Infrastructure Documentation.
• Preambles: Initial context can be given to the LLM through the options.preamble parameter, helping guide its responses better. For details on using a preamble, refer to the Offered Pretrained Foundational Models in Generative AI in the Oracle Cloud Infrastructure Documentation.
• Image processing: You can provide an image (as a file.File object) using the options.image parameter. You can ask questions about the image and receive responses that can include advanced image captions, a detailed description of the image, information about charts and graphs in the image, and so on.

Conclusion

These models and capabilities enrich the functionality of the SuiteScript generative AI APIs, enabling more flexible and informed responses from the LLM according to user-defined parameters.

Key Takeaways

• The table outlines models supported by SuiteScript generative AI APIs.
• RAG enhances responses by incorporating additional document context.
• Preambles help set context for the LLM's generation process.
• Image processing allows for direct inquiries about images.