The AI model has stopped working properly

Support Help
1

Where does the bug appear (feature/product)?

Cursor IDE

Describe the Bug

The AI ​​model has stopped working properly.
Sometimes the model is capable of handling even the most complex projects with a huge number of lines, but in the last few days I decided to create a new project and wrote the specifications. And I can’t seem to get it implemented. The AI ​​assistant tells me the project is 100% complete, although it’s only 5% complete at most. And I can’t convince it to correct the errors because it keeps doing the opposite. The assistants are behaving very strangely.

Steps to Reproduce

The AI ​​model has stopped working properly.
Sometimes the model is capable of handling even the most complex projects with a huge number of lines, but in the last few days I decided to create a new project and wrote the specifications. And I can’t seem to get it implemented. The AI ​​assistant tells me the project is 100% complete, although it’s only 5% complete at most. And I can’t convince it to correct the errors because it keeps doing the opposite. The assistants are behaving very strangely.

Expected Behavior

…

Operating System

Linux

Current Cursor Version (Menu → About Cursor → Copy)

…

Does this stop you from using Cursor

Yes - Cursor is unusable

2

I think there’s something wrong on the server end for the last 30 minutes or so. Cursor won’t process any prompts. It keeps asking how it can help instead of doing the task. Maybe this is related?

3

Hey, thanks for the report.

I see that @CyberDragon also noted AI issues in the last 30 minutes. This might be a temporary server-side problem.

To investigate, we need a bit more info:

  1. Cursor version: Help > About Cursor > Copy
  2. Which model is used (Claude Sonnet 4.5, GPT-5, etc.)?
  3. Request IDs: in chat, click the context menu (top-right) > Copy Request ID for the problematic requests
  4. Network diagnostics: Cursor Settings > Network > Run Diagnostics - what’s the result?
  5. Can you share a screenshot or a concrete example of the prompt and the AI’s reply showing the odd behavior?

Let me know if the issue is still happening or if it resolved itself.

4

It resolved itself just now. Cheers!

1 Like
5

What model were you using?

That happens to me occasionally as well… It’ll show “Planning next moves…” perpetually without actually executing it. You can try switching the model, as it’s not necessarily cursor itself being “unavailable” or “not working;” it could be the specific provider of the model you have selected.

In the future, you could also try restarting.

Happy developing!

1 Like
6

Of course, I use all the models one by one, hoping to find one that works. In general, everything works perfectly in one project, but in another, everything is the opposite. Every model does everything incorrectly, and I can’t get it to perform its job properly. It even estimates its effectiveness at 20%, although in reality, it’s no more than 5%.
I’m sure this is due to artificial limitations. That is, projects like those that are commercial or convenient for market analysis or creating a full-fledged trading bot won’t achieve the desired results.

7

This topic was automatically closed 22 days after the last reply. New replies are no longer allowed.