bug: GPT-4 doesn't support vision & file search in Jan

[imtuyethan]

• I have searched the existing issues

Current behavior

https://discord.com/channels/1107178041848909847/1279700963439022090
Users are unable to use file-related features in Jan, including GPT-4 Vision capabilities (image analysis) and document upload for RAG (Retrieval-Augmented Generation), despite having added their OpenAI API key and selecting the GPT-4 model. These functionalities should be available but appear to be non-functional.

Minimum reproduction step

1. Add OpenAI API key to Jan
2. Select GPT-4 model
3. Attempt to send an image for analysis
4. Attempt to upload a document (e.g., PDF) for RAG

Expected behavior

Jan should be able to:

• Process images using GPT-4's vision capabilities
• Allow document uploads for RAG
• Provide responses or allow for questions about the uploaded content

Screenshots / Logs

Jan version

v0.5.3

In which operating systems have you tested?

Operating System: Pop!_OS 22.04
KDE Plasma Version: 5.24.7
KDE Frameworks Version: 5.92.0
Qt Version: 5.15.3
Kernel Version: 6.9.3-76060903-generic (64-bit)
Graphics Platform: X11
Processors: 16 × 13th Gen Intel® Core™ i5-13500H
Memory: 15.3 GiB of RAM
Graphics Processor: Mesa Intel® Graphics

Btw, I have a dedicated Nvidia 4060 as well.

[freelerobot]

related #3505

[imtuyethan]

Close this ticket as dup?

[freelerobot]

discussions: Remote API Extension #3505

Leave it open bc:

1. We haven't really impl RAG yet
2. We don't know RAG will work with remote endpoints

[imtuyethan]

OpenAI GPT-4o model’s vision capability is enabled: