Local inference of LLMs is essentially free. There doesn’t exist a sufficiently deep pool of experts of all knowledge spaces who are also willing to work for free, who can be trivially identified, contacted, and scheduled.