You should never conflate products with models, it leads to invalid comparisons between models and products. Which for some reason has been the norm. How many times have you heard someone ask "How much worse is Llama than ChatGPT?"?