About Intecca

Intecca answers one question well: which open LLMs can your computer actually run? Pick your GPU, Mac or RAM and it computes the memory each model needs, the best quantisation that fits, how much context you have room for, and a rough speed — for 30 current models across 35 hardware configurations.

It exists because the usual answers are bad. Listicles go stale within weeks of a new release and rarely tell you what fits your card. General chat assistants don’t track this month’s models and won’t do the per-rig memory math reliably. The arithmetic isn’t hard — it’s just nobody’s job to keep it current and personal. So we did.

Everything is computed transparently from published model architectures and hardware specs; the formulas and their limits are on the methodology page. The calculator runs entirely in your browser — nothing about your setup is sent anywhere.

How it stays useful

The model list is versioned and updated as new weights land, because this space moves weekly. The estimates are deliberately honest: clear verdicts, stated assumptions, and no pretending a number is exact when it’s a careful approximation.

How it’s funded

Intecca is free. Some links to hardware are affiliate links, which may earn a small commission if you buy — they never change which model or card we recommend. See privacy for details.