I have spent the past few months working on a project I call phntm-one. Its a Raspberry Pi5 running a 4B LLM model installed into a 10 inch touchscreen. I created an app for the LLM to run within. You can chat with it, talk to it and it talks back. The context I was able to get is around 9K out of the 4B model. I also added RAG and persistent memory. The appliance will even repair itself and learn from the fixes for future use and avoidance. Please check out my website for a full list of features and what it can do. I have spent a lot of time building this and am proud of my end product. I still have a lot of work to do but the device is just about fully autonomous.