alt Hacker News>> My guess would be that the training data differs just enough for it to have an impact.
It's because performance degrades over longer conversations, which decreases the chance that the same conversation will result in a solution, and increases the chance that a new one will. I suspect you would get the same result even if you didn't switch to a different model.
So not really, certainly models degrade by some degree on context retrieval. However, in Cursor you can just change the model used for the exchange, it still has the same long context. You'll see the different model strengths and weaknesses contrasted.
They just have different strengths and weaknesses.