Benchmarks for the AI you run yourself.
Pick the context you need and the memory you have, and see which open models fit your hardware and how well they actually perform.
Pick the context you need and the memory you have, and see which open models fit your hardware and how well they actually perform.