I know this has been asked several times, but I cannot see any solutions. It worked for several weeks perfectly now and then suddenly it stopped working. I have a pro account fyi.
Is there a cache folder I can try to delete, I tried restarting, tried deleting the Index and starting over but it just keeps on failing now on folders that I know 100% worked. Is there anything I can try to fix this?
I don’t think it’s anything you’re doing wrong Johnsters. There is an ongoing issue with OpenAI’s Embeddings, which is resulting in issues with Codebase Indexes. See here for updates: https://status.cursor.com/
I am also experiencing the same issue, is there an expected date for when is going to be fixed ?
I see that issues like this have happened before
Hey, this should be fixed now!
We rely on OpenAI for this part of Cursor’s functionality, which has been having issues for the majority of the day, but a fix is now rolled out and codebase indexing should be back to full functionality.
Thanks for the explanation and the update !
I can confirm is working now 
For some reason I still get the handshake failed error, I changed the model of the indexing and still the same for some reason
Hey, is everything resolved now? We had a few users who may have had some teething issues here, but we believe everything should be working as normal now.
Hello, Not for me at least.
I see it working right now !
Yes, the model TextEmbeddingsLarge3 now works but the other model VoyageCode2 does not work sadly
I learned that VoyageCode2 is better with coding so I hope there will be a fix for it
Hi. I’ve started experiencing issues from about an hour ago. Indexing is failing and agent is not following instructions given. I have to provide an instruction in the chat to the letter even though I have it in my instructions (rule of ai).
same here, it seems the main server is down or not responding. I get “Handshake failed” when trying to index. For some reason it was able to delete the index, or at least it said it did.
Hey @snowpearsor and @non-npc, we had some downtime a few hours ago that affected all our services, but we should be back up and running now!
ry disabling HTTP/2 by setting:
cursor.general.disableHttp2
It might not be the root cause, but this worked for me. Worth a shot! 
