Recent discussions among AI enthusiasts indicate that OpenAI's GPT-4o models are outperforming Anthropic's Claude 3.5 Sonnet in various tasks. Users have reported that while Claude 3.5 Sonnet can be effective for certain coding projects, it falls short in reasoning and general question-answering compared to GPT-4o. The GPT-o4 model also shows better performance in fixing typos. Furthermore, some users who switched to Claude have reverted to using ChatGPT due to dissatisfaction with Claude's rate limits. Overall, OpenAI's models are currently seen as superior in performance.