AI research product

Local LLM Lab

A measured guide to every local AI model installed on my Apple Silicon workstation.

Local LLM Lab field guide showing the current installed-model inventory and measured comparison totals.
Local LLM Lab field guide showing the current installed-model inventory and measured comparison totals. Download the editable Excalidraw source.

What it is

Local LLM Lab turns a large collection of local model builds and benchmark evidence into a practical field guide. It shows what is installed now, what each build is good at, which input types and tool features it supports, and how quality, speed, memory use, and context length compare on the same machine.

Python scripts inspect the live Hugging Face, Ollama, MLX, and Apple model inventories, join them to checked-in benchmark results, and generate a static React guide. Exact model revisions stay separate because the runtime and quantisation can materially change speed, memory use, and output quality.

Evidence & links

Related reading