Comment by a_vanderbilt

My experiences have been the same, unfortunately. It can do simple tasks, but for anything requiring indirect reasoning or completion of partial content from media (think finishing sonnets as a training content test) Claude just falls flat. Honestly, I'm not sure what makes Claude so "meh". Not to mention having to fill out a Google Doc for API usage? Weird.