# Very poor performance using model weights **URL:** \ **Category:** 🤝 Community Help\ **Created:** [April 23, 2025, 9:27pm UTC](https://discuss.roboflow.com/t/very-poor-performance-using-model-weights/9585 "2025-04-23T21:27:48Z")\ **Posts on this page:** 8\ **Page:** 1
**Author:** ![EthanRhode](https://avatars.discourse-cdn.com/v4/letter/e/7feea3/32.png) [@EthanRhode](https://discuss.roboflow.com/u/EthanRhode)\ **Post date:** [April 23, 2025, 9:27pm UTC](https://discuss.roboflow.com/t/very-poor-performance-using-model-weights/9585/1 "2025-04-23T21:27:48Z")
When downloading model weights (Not using this account, if anyone from roboflow is wondering how I am downloading weights on Growth plan) the performance is much worse. I don’t have comparison pictures (Ran out of credits, which is why I am using kaggle for YOLO11 inference), but it is significantly worse. I attached one of the images I inferred using kaggle. Using Roboflow inference, It would detect all numbers besides maybe 1 or 2. Using kaggle, well you can see the difference. Does roboflow have something that makes it better at inference? ![inferred_image](https://canada1.discourse-cdn.com/flex029/uploads/roboflow1/original/2X/7/7b20ae4e133236d77fcb7a19832839926eb9a973.jpeg) ---
**Author:** ![Jacob\_Witt](https://yyz1.discourse-cdn.com/flex029/user_avatar/discuss.roboflow.com/jacob_witt/32/347_2.png) [@Jacob\_Witt](https://discuss.roboflow.com/u/Jacob_Witt)\ **Post date:** [April 23, 2025, 9:52pm UTC](https://discuss.roboflow.com/t/very-poor-performance-using-model-weights/9585/2 "2025-04-23T21:52:14Z")
Hi there @EthanRhode - my gut is you have an error in your deployment pipeline somewhere. We’ve spent a ton of time ironing out common errors in our inference server, which is likely why you’re seeing good results there. ---
**Author:** ![EthanRhode](https://avatars.discourse-cdn.com/v4/letter/e/7feea3/32.png) [@EthanRhode](https://discuss.roboflow.com/u/EthanRhode)\ **Post date:** [April 23, 2025, 11:59pm UTC](https://discuss.roboflow.com/t/very-poor-performance-using-model-weights/9585/3 "2025-04-23T23:59:22Z")
I just tested it using an RF-DETR model and the same thing happened with reduced performance on inference. I created a new notebook in kaggle and copy pasted the RF-DETR inference script and just changed the paths for the model and images. Was there something else I was supposed to do? ---
**Author:** ![Jacob\_Witt](https://yyz1.discourse-cdn.com/flex029/user_avatar/discuss.roboflow.com/jacob_witt/32/347_2.png) [@Jacob\_Witt](https://discuss.roboflow.com/u/Jacob_Witt)\ **Post date:** [April 24, 2025, 11:57am UTC](https://discuss.roboflow.com/t/very-poor-performance-using-model-weights/9585/4 "2025-04-24T11:57:13Z")
Sorry - want to make sure I’m following. You’re seeing better results for yolov11 on inference, but worse results for RF-DETR? ---
**Author:** ![EthanRhode](https://avatars.discourse-cdn.com/v4/letter/e/7feea3/32.png) [@EthanRhode](https://discuss.roboflow.com/u/EthanRhode)\ **Post date:** [April 24, 2025, 1:13pm UTC](https://discuss.roboflow.com/t/very-poor-performance-using-model-weights/9585/5 "2025-04-24T13:13:19Z")
No, I am seeing worse results for both running inference on kaggle. ---
**Author:** ![Jacob\_Witt](https://yyz1.discourse-cdn.com/flex029/user_avatar/discuss.roboflow.com/jacob_witt/32/347_2.png) [@Jacob\_Witt](https://discuss.roboflow.com/u/Jacob_Witt)\ **Post date:** [April 24, 2025, 1:40pm UTC](https://discuss.roboflow.com/t/very-poor-performance-using-model-weights/9585/6 "2025-04-24T13:40:47Z")
My advice is to just self-host our inference server – I’m not able to troubleshoot custom deployment pipelines. ---
**Author:** ![EthanRhode](https://avatars.discourse-cdn.com/v4/letter/e/7feea3/32.png) [@EthanRhode](https://discuss.roboflow.com/u/EthanRhode)\ **Post date:** [April 24, 2025, 2:08pm UTC](https://discuss.roboflow.com/t/very-poor-performance-using-model-weights/9585/7 "2025-04-24T14:08:49Z")
Okay, Thank you. I assumed you wouldn’t be able to but I figured I would try my luck. ---
**Author:** ![system](https://canada1.discourse-cdn.com/flex029/uploads/roboflow1/original/1X/bb6edbfa56d0cf2ee603f061e9e09103502cc8cc.png) [@system](https://discuss.roboflow.com/u/system)\ **Post date:** [May 15, 2025, 2:09pm UTC](https://discuss.roboflow.com/t/very-poor-performance-using-model-weights/9585/8 "2025-05-15T14:09:38Z")
This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.