The arrival of the 86B model marks a crucial change in the landscape of large language models. Its impressively smaller footprint compared to powerhouses like GPT-4 doesn't necessarily suggest weakness, 86blegal.com