Add multi-modality features

[jorgeantonio21]

Feature Request

Currently, we only support text based models for the chats. The following models are the most important to integrate atm:

• Llama3.2 Vision models.
• Flux (text-to-image) models.
• Whisper (voice-to-text), or any alternative.

Moreover, each chat session should have a multi-modality feature. For example, if a user interacts with a specific character, it can request it to analyze and give feedback on images, etc.