Loading
Upcoming Mandatory Changes to Public Key Infrastructure (PKI)Read More
Agentforce and Einstein Generative AI
Table of Contents
Filter by (0)Add
Select Filters

          No results
          No results
          Here are some search tips

          Check the spelling of your keywords.
          Use more general search terms.
          Select fewer filters to broaden your search.

            Search all of Salesforce Help
            Search all of Salesforce Help
            Grounding with File Inputs

            You are here:

            1. Salesforce Help
            2. Docs
            3. Quickstart Your Einstein Generative AI Solution

            Grounding with File Inputs

            With file inputs, you can ground your prompt templates with richer context to achieve more accurate and relevant responses. Incorporate images and PDF files into your prompt template. Image and PDF inputs are useful for scenarios where your prompt must process and analyze visual information.

            Required Editions

            Available in: Lightning Experience
            Available in: Enterprise, Performance, and Unlimited Editions with the Einstein for Platform, or Einstein or Agentforce for Sales or Service add-on, or Agentforce Foundations

            These file types are supported.

            • .pdf
            • .png
            • .jpeg
            Note
            Note
            • Not all models support file inputs. Before adding images or PDF files, review your model’s file size and quantity requirements. Unsupported files are excluded from the request. The excluded files are listed in the resolution panel after processing. To learn more, see Prompt Builder Limits.
            • During prompt template execution, a JPEG, PNG, or PDF file is uploaded as a binary file in the prompt request. Einstein Trust layer protections are not applied to the contents of the uploaded binary file. To apply Einstein Trust layer protections for a PDF file, in the prompt template, first extract the applicable content of the PDF (in text format) into the prompt, and then send the prompt to the Large Language Model.

            After an image or PDF is selected, the file and its metadata are associated with the prompt template. Metadata here refers to the metadata of the file record, as all used files are uploaded into your org as contentDocuments and, in turn, have their own record page. If the metadata of an image is needed in the prompt template, you can select it by clicking the Resources field in the Prompt section and use it as a merge field.

             
            Loading
            Salesforce Help | Article