Tools
Pure client-side, no sign-up, shareable by URL. Each one feeds a kit.
LLM VRAM calculator
Pick a model, quantization and context; see which GPUs and Macs fit and roughly how fast they'd run it.
Which machine should I buy to run AI locally?
Three questions. One ranked answer, learned from what people in your situation actually picked.
Coming next: kit power-draw and electricity cost planner; $/GB VRAM table. Request one.