its just the press keep pushing fake news,
comparing the results of actual work with cursor vs. z.ai and trae.ai
grok/gpt/others wins big time, glm , deepseek are constantly failing when used in real world scenarios
I havent noticed that big a gap when working with different models