I'd like the ability to set a default number of LLM's. Right now the same three LLM's are always chosen and it takes a few steps to have to remove them and add the ones that I want.

Your app has two areas that require improvement. Firstly, despite initiating a new chat, it continues to retrieve information from previous chats, which poses a privacy and security risk. It's concerning that prior chat data persists in new conversations.

Secondly, the chat defaults to three models that are often unhelpful. For instance, if I choose GPT-4o, I expect it to remain as GPT-4o during subsequent interactions rather than reverting back to default settings.

I'd like the system to remember, per conversation, the selected model(s) used. It tends to reset to the "default 3" each time which is unnecessary and wasteful of credits.

When using the app, mostly for text content generation, I have to manually set the parameters each time I do a new prompt.

Which leads to errors manytimes and credits used for nothing because if I forget to set the parameters (like the lang to FRENCH for exemple), I have to relaunch the request.

Red area on my attached screenshots should be saved somehow, or put by default when I re-use the app. Sometimes I forget to select the correct AI, sometimes it's the language...

This would be a really nice feature. Beeing french myself, and having english set by default for content generated, even if FRENCH is set for the UI, is complicating the work.

Regards

Furthermore , implementation of own profile to take into consideration and avoid wasting credits.

In multiDoc chat I want to be able to store my AI engine selection. Eg I always get gemini and I want to have Clause of ChatGpt as default model

When moving between AI tools the settings pane that allows us to change the AI model, toggle mix AI models, view credit estimate etc. collapses. It would be so useful if this could be set to always be shown by default.

I know there is currently a request to have default settings but I think this would be a separate request to the one I'm making.