Yes, local LLMs are ready to ease the compute strain