Qwen 3.8 performs fast prompts, generating poetry and Houdini VEX code for particle deletion requests

I'm just starting my journey here into local LLM. Since I work in VFX I've got a 5090 and 256GB of sys mem. Installed LM studio and downloaded Qwen 3.8 27B since that very comfortably fits in VRAM. Set it to low...