We need a consistent way to tell level of service for LLM services. Without they can play with us and steal money by providing different/quantized models than we paid for. I believe this may be a good next AI company if they will be reliable.