The A.I. Daily
Seeking models for 12GB VRAM GPUs
A Reddit LocalLLaMA thread asks 12GB VRAM 3080 users what models work well locally; Qwen 3.6 35B 3A was cited as common, and the OP seeks models that leverage the extra 4GB for agent-style personal assistant workflows.