The recently announced GPT-4o model has 17 samples actively listed on the official announcement page. I compared its closest competitors (also listed as competitors by OpenAi). I compared it with AIs that can do the same prompts and labels.
When I compare the data, it’s clear that GPT has a clear lead. While it may not be as versatile as GPT, it excels in its field. The “Humanoids” are too focused on the answer part and the output is sometimes unnecessarily inaccurate. However, there’s still plenty of room for improvement! I will share the tests under this title.