We Ran the exact same reasoning check on Llama 3.1 70B which it did not answer accurately, but in all kinds of other assessments, the model excelled. Also, Meta AI also offers picture generation which happens to be good. Pi begins the dialogue alone and attempts to glean in each https://poe-chat.online/