Testing 20+ local llms to see how good they are at understanding large contexts
People keep saying OpenAI has no moat. I think they're wrong - the answer is apps.