Local AI lab

I’ve been setting up local model workflows across a small Mac and an Ubuntu workstation. The Mac gives me a tight memory budget to work with; the workstation gives me room to explore larger quantized models and GPU constraints.

Ollama and installed models are verified on these setups, but that is a starting point, not a benchmark. I haven’t established inference speed or task-quality results, and a cloud-backed model entry is kept separate from the local ones.