The arrival of the 86B model signals a crucial change in the field of large language models. Its impressively smaller size compared to giants like GPT-4 isn't necessarily imply weakness, but rather 86blegal.com