DeepSeek Flash 4.1, the worst version of DeepSeek ever for those who enjoy learning and studying.

The Flash 4.1 version is perhaps the worst release DeepSeek has had since the launch of R1 Beta. In terms of speed, it stands out, and it is also lightweight! However, for studying or truly delving deep into any subject, this model is an unprecedented disaster in DeepSeek's history over the last two years. It is a completely useless model when it comes to explaining any topic for those who genuinely enjoy learning. The model fails to provide depth on anything; explanations are reduced to one-line bullet points, even when explicitly asking the model to elaborate. This truly represents a total step backward. It is incredible how the DeepSeek team still claims that version 4.1 is better than the Pro 4.0 version. That is a complete lie. Model 4.1 is not even better than the previous Flash 4.0 version; Flash 4.0 could explain and delve deeper into topics far more effectively than this 4.1 iteration. I have been using DeepSeek for two years. I only stopped using it during version 3.2, which was also a disaster, particularly regarding critical details and nuances, as well as its excessive flattery. However, I believe this Flash 4.1 version manages to be even worse on certain levels. For those who enjoy philosophy or truly profound subjects, this is the worst version DeepSeek has ever released. DeepSeek always stood out for being excellent for those who liked to study and learn about more advanced concepts... and now, though it pains me to say it, it is garbage. It is entirely unviable for me. I don't even understand the real reason behind the DeepSeek team doing this... it's a complete own goal!

添加评论
点赞收藏
点踩分享查看原文
评论
?
参与讨论