[Feature Request]: include attachments support (PDFs, images, etc)

[younes-io]

Problem Description

I cannot use NextChat to "chat with my documents" or just take a screenshot of some code snippet or some mathematical formula and then use that in my chat with the LLM

Solution Description

I want to be able to:

• upload documents to my conversation (PDF, Word, raw text files like CSV, PowerPoint, Images) and chat with them (this is the priority for me)
• define a Mask/Assistant (just like in the OpenAI playground) that is connected to a knowledge base (one or a bunch of docs) and then chat with it whenever I want to => it will be a specialized Mask, that answers based on my own knowledge base, not just relying on the training data of the LLM with a system prompt

Alternatives Considered

ChatGPT (for the first item)
OpenAI's Assistants (for both items)
ChatKit

Additional Context

No response

[Dean-YZG]

Thank you for your feedback, we have planed to build data storage capacity for the future to upload files, but you know, building data storage capacity is not only technically complex, but more importantly, a data security issue. Therefore, we will build this capacity after developing a reliable technical solution

[younes-io]

@Dean-YZG even for local data storage?

[Issues-translate-bot]

Bot detected the issue body's language is not English, translate it automatically.

---

@Dean-YZG even for local data storage?

[BMurdochUK]

But the Gemini-pro-vision model allows images to be uploaded via the NextWeb app. So guessing it persists temporarily to local data storage? So why cant same feature be rolled out to ChatGPT model?

#4446

[younes-io]

@Dean-YZG even for local data storage?

@Dean-YZG could you please comment my question or someone from the team? =) Thank you