Degradation of service

Describe the Bug

For weeks now, regardless of the model I use, the application has experienced a dramatic degradation in service. It constantly freezes, makes a lot of mistakes, and doesn’t even follow detailed prompts. I’ll spend hours correcting an error, only for it to revert to the same mistake later on. I’m on the Pro plan and even pay an additional monthly fee. At first, it was worth it, but now it’s making too many errors, crashing, getting stuck, and providing slow responses. It also doesn’t show the files being modified in the IDE.

Steps to Reproduce

For example, I’m having a problem with the implementation of an email sender. I spent about 4 hours testing an implementation with a mailer.

Expected Behavior

Please fix the behavior. It’s constantly getting stuck and is no longer as accurate in its reviews and proposed corrections. This started about 3 or 4 weeks ago, when I first installed the updates.

Operating System

Windows 10/11

Current Cursor Version (Menu → About Cursor → Copy)

Version: 1.4.5
VsCode: 1.99.3
Node.js 20.19.1

Additional Information

I’m working on an application with Java, Quarkus, Panache, and a frontend using React and Tailwind. The architecture is a business layer.

Does this stop you from using Cursor

Yes - Cursor is unusable

2 Likes

hi @Sergio_Zamora and thank you for your bug report.

Could you clarify which model are you using?

We do not see any systematic degradation of services but I would be happy to review cause with you.

Please check Settings > Network > Run Diagnostic and report any errors shown there.

Could you also add a Request ID with privacy disabled so we can look into the details? Cursor – Getting a Request ID

This is true, and I think it’s intentional from Cursor. I feel the same, I used to accomplish twice as much in the same time before. Now needs more corrections, calls, I think they are doing this intentionally to squeeze more money. It was way better a month ago.

@Kaushik_Medikonda thank you for your feedback but note that your insinuation we may be doing this on purpose is not correct. I am happy to review detailed bug reports and check for errors.