Here is my experience, I've been using LM Studio with RTX 5080 and 64GB RAM using Windows 11. I'm very happy with LM Studio except the speed. I installed Windows WSL and compiled Llama.cpp. After playing with Gemma 4 26B Q8 and Qwen 3 Coder Next unsloth Q4 with Llama.cpp, I'm getting double the speed compared to LM Studio. I wish LM Studio provided the same speed, but unfortunately, it doesn’t.
[link] [comments]