[Enhancement]: disable OCR in file upload

[marburps]

What features would you like to see added?

I am unable to configure LibreChat (running in docker) to prevent OCR from happening in any file upload, even plain text or for file search. the upload always fails because of OCR wanting to happen. There should be no need to perform OCR on text files.

More details

I tried every possible configuration in librechat.yaml or .env , even created a costum agent, but I cannot get around OCR happening on file upload.
Basically this means the complete file upload is useless if you don't have a subscription on any of the 3 offered OCR providers.
Should this possibility however exist, then it is so well hidden that it is not possible to find it into the documentation with any reasonable effort.

Which components are impacted by your request?

General

Pictures

No response

Code of Conduct

• I agree to follow this project's Code of Conduct

[danny-avila]

endpoints:
  agents:
    capabilities:
    - "execute_code"
    - "file_search"
    - "actions"
    - "tools"
    - "artifacts"
    # - "ocr" # Omit OCR
    - "chain"
    - "web_search"

https://www.librechat.ai/docs/configuration/librechat_yaml/object_structure/agents#capabilities

[chillwalker]

I understand this is the configuration for an agent. How do I achieve this in the regular chat?

[delastone1-gif]

Can we get an answer about this , outside the agent model ?