CPU
CPU-Only LLMs: What Actually Works
A practical guide to running LLMs on CPU only — no GPU required. Covers what models work on laptops and desktops, a budget dual Xeon server build for 70B models, and when CPU-only makes sense.
Best Models Under 3B Parameters: Small LLMs That Actually Work
The best sub-3B language models for laptops, old GPUs, Raspberry Pi, and phones. Covers what works, what doesn't, and which tiny model to pick for your hardware.