Post #497034

@User47@vmst.io Just install ollama. Then you can use it to pull and run all kinds of models. I just generally use Deepseek-R1 as it's a decent reasoning model and a little more focused for my use cases. You just have to find the right parameter size for your hardware that gives you a reasonable compromise of complexity and utility vs performance. I generally use the 8b model on a mobile 3060 in my laptop.