R1 and Qwer Support and Cursor Tips?

I have been using Cursor for a while and I am very pleased. It can do very good work compared to many other AIs and tools. I also complete my work very quickly and it saves time. This is a fact and Cursor is quite successful compared to other tools. However, I was wondering, Qwen and R1 are free. It is also recommended and wondered about a lot. However, no steps have been taken regarding Qwen in Cursor yet. When I try R1, I realize that it repeats itself a lot. It works very fast and well until it gives an error. However, when it gets stuck at some point, it repeats the same thing a lot. That’s why Claud still seems to work much more efficiently. It finds different perspectives and can try different things. Sometimes it breaks a lot because of this, but at least it brings perspective. I couldn’t see this in R1 for some reason. Maybe I am using it wrong. Or R1 could not be fully integrated into Cursor, I don’t know. Anyway, when I try to find a result, for some reason it extends the process unnecessarily under the name of thinking and sometimes I realize that it does not fully understand. I think there is a problem especially in the Turkish language. That’s why I was wondering, will there be a step towards developing Qwen or R1 right now. It seems like a much more efficient, cheaper and higher quality work will come out of here. While Claud works perfectly integrated, R1 is not like this, I think it will be much more efficient if it can be integrated. I’m also curious about how Qwen will be.

By the way, I sometimes have problems with the codebase structure. No matter what we do, artificial intelligence forgets the code structure. I understand that. That’s why I was wondering, is it possible to do something different and remember the code structure? Is there anyone who can create a more logical and dominant structure with Cursor.rules or different solutions? If anyone can give me an idea about this, I would be happy. I am currently working mainly with Laravel and Flutter. Is there anything you can suggest specifically for cursor tips?

Thanks and sorry for the long post, I am just excited and hopeful about the developments. I wonder how I can use it in a more efficient and better way ^^

Hey, both DeepSeek v3 and R1 are not available in Cursor, you can enable them in the Cursor model settings page.

While Qwen is not available, we are always looking for new and better models to add, and will add them whenever they cross some internal thresholds on performance improvements and speed.

Thanks for the answer. I tried R1, yes. It works very fast and well. However, as I said, there are points where it gets stuck, and it extends too much under the name Think. I think it can be integrated into Cursor in a better way. I don’t think it works with full performance. It will be great if improvements are made. Especially if it gets stuck at some points, it starts to recur and tries to do the same process over and over again and doesn’t suggest new solutions. This was a bit annoying, so Claud is still more logical.

They recommend Qwen code a lot and as far as I know, it came out free. It would be good if it could be added. I can do some tests there. The 70 thousand version was published, if I’m not mistaken, and it was said that it had results close to Claude. I don’t know how true it is, but it would be nice to try it if it is successful.

I find your work very successful. I hope Cursor continues to develop even better. I especially believe that one day we will see a structure that completely solves the codebase and progresses logically without forgetting. I think there is still no company that can do this completely. Although Cursor solves the codebase, it starts to forget as the process progresses. We have to remind it from time to time and continue. I believe that one day we will be able to do transactions by completely remembering it ^^

1 Like