Problem Statement
I use local LLMs for converting hand-written and printed documents to text. That works very well with vision capable models like Gemma4. However, those models require the input as an image while many scanners/MFP printers produce PDF output with the image embedded.
Feature Idea
When attaching a PDF, there could be a third option besides "Embed" and "Include in chat" - "Convert to image".
Problem Statement
I use local LLMs for converting hand-written and printed documents to text. That works very well with vision capable models like Gemma4. However, those models require the input as an image while many scanners/MFP printers produce PDF output with the image embedded.
Feature Idea
When attaching a PDF, there could be a third option besides "Embed" and "Include in chat" - "Convert to image".