clear_view·昨年·議論deepseek-r1:14b/mistral-small:24b/qwen2.5-coder:14b fit 16GB VRAM with fast generation. 32b versions bleed into RAM and take a serious performance hit but still usable.
clear_view·昨年·議論32B don't fully fit 16GB of VRAM. Still fine for higher quality answers, worth the extra wait in some cases.