I have searched the existing issues and discussions.
Problem Description
Currently, open webui will use OCR anyway when uploading documents.
However, for vision-supported models, just convey the images of documents may be better, especially when there're some pictures in documents that OCR cannot handled.
Desired Solution you'd like
Provide an option in documents, enabling open webui to convert documents, such as PDFs or docx, to images, and pass images to models instead of text results when documents are uploaded for vision supported models. For models that do not support vision, use the original OCR solution.
Alternatives Considered
No response
Additional Context
No response
Originally created by @funnycups on GitHub (Mar 10, 2025).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/11479
### Check Existing Issues
- [x] I have searched the existing issues and discussions.
### Problem Description
Currently, open webui will use OCR anyway when uploading documents.
However, for vision-supported models, just convey the images of documents may be better, especially when there're some pictures in documents that OCR cannot handled.
### Desired Solution you'd like
Provide an option in documents, enabling open webui to convert documents, such as PDFs or docx, to images, and pass images to models instead of text results when documents are uploaded for vision supported models. For models that do not support vision, use the original OCR solution.
### Alternatives Considered
_No response_
### Additional Context
_No response_
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Originally created by @funnycups on GitHub (Mar 10, 2025).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/11479
Check Existing Issues
Problem Description
Currently, open webui will use OCR anyway when uploading documents.
However, for vision-supported models, just convey the images of documents may be better, especially when there're some pictures in documents that OCR cannot handled.
Desired Solution you'd like
Provide an option in documents, enabling open webui to convert documents, such as PDFs or docx, to images, and pass images to models instead of text results when documents are uploaded for vision supported models. For models that do not support vision, use the original OCR solution.
Alternatives Considered
No response
Additional Context
No response