Crazy they didn’t mention this new language model. I wonder if it’s multimodal or if it’s multiple models