OP here. I spent the last 3 months building this because I wanted a cleaner UI for the Mistral ecosystem.

The Stack: React, Node.js, Mongo, Mistral API. The Problem: Most wrappers are just 'stateless' chat. I wanted persistent memory that felt like an OS. Why Mistral?: I found 'Mistral Large' to have better reasoning for my specific coding tasks without the 'verbosity' of GPT-4.

Deeply appreciate any feedback on the latency or the memory retrieval logic

https://queueai.app/