Some Cursor users suspect GPT behavior changed this week
GPT inside Cursor may have felt noticeably worse to use during the past week. There is no confirmed cause or official change mentioned.
The available information is a broad complaint about model behavior, not a detailed test with examples or numbers. The main point is : the change could be real, or it could come from one person’s , prompts, or expectations.
Key points
- GPT behavior in Cursor was reported as noticeably different this week.
- No official explanation or confirmed model change is included.
- No concrete examples, tests, or numbers are provided.
- The issue may be a real change or a flow problem.
- For important work, compare the same prompt across models before deciding.