Observed arrival · 2026-10-19
RigFit estimates which local models fit your hardware
A public checker compares local AI models with GPUs, Macs, and mini PCs across Q4, Q8, and FP16 quantizations.
○Open
⊠Login
$Paid
†Ads
✦Pretty
●Pro
◎Niche
◉Human
⚑Risk
ƒJS
- For
- People choosing hardware for local language models
- Worth noticing
- The value ranking counts Q4 model fits per $100 of list price, but explicitly excludes memory bandwidth from its score.
Field notes
The estimate uses a matching GGUF file size when one is stored; otherwise, it calculates from parameter count and bytes per weight. Its stated 20% overhead allowance is not intended to cover very long contexts, and the site gives separate usable-memory assumptions for GPUs, Macs, and Strix Halo mini PCs. A value ranking counts Q4 fits per $100 of list price but leaves bandwidth out of the comparison.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
○OpenPublic substance visible
$PaidCommerce or pricing visible
✦PrettyNotable craft visible
●ProPolished or operationally mature
◎NicheUnusually specific use
One card from the complete issue