Discussion: when does a local LLM rig pay back?

A Hacker News post examines when a local LLM rig pays for itself by modeling hardware, model, and daily token usage to compare local run costs versus renting the same model; the author asks for improvement ideas.