choose model

[patrikhalili]

Feature idea: Choose local GGUF model from Downloads and run it directly

I would like Odysseus to have a simple local model workflow like this:

1. User clicks “Choose model” or “Import model”.
2. User selects a .gguf model from Downloads or any folder.
3. Odysseus checks if the file is a valid GGUF model.
4. Odysseus copies the model into its own local models folder.
5. Odysseus registers the model automatically.
6. The model appears in the model selector.
7. User clicks it and it runs locally with llama.cpp / llama-server.

The goal is: no terminal commands, no manual editing, no complicated setup.

It should also check model size against available RAM and warn the user if the model is too large for the computer.

This would make Odysseus much easier for normal users:

• download a GGUF model
• open Odysseus
• choose the model
• start chatting locally