Do we even need Anthropic or OpenAI's top models, or can we get away with a smaller local model? Sure, it might be slower, ...
My homelab actually pays off now.
Holy*****, where has this been all year?