Observed arrival · 2026-11-12
Can I run this model?
A calculator for estimating whether an open language model can meet memory, speed, and first-token targets on selected hardware.
○Open
⊠Login
$Paid
†Ads
✦Pretty
●Pro
◎Niche
◉Human
⚑Risk
ƒJS
- For
- People sizing hardware for open language models
- Worth noticing
- It models mixture-of-experts weights separately from per-token cache, and warns that speed estimates are usually within ±30%.
Field notes
Users can enter or select a model, choose a quantization, describe hardware, and set concurrent-request, prompt/reply-length, speed, and first-token targets. The page says it tests catalog models at progressively smaller formats and checks GPU splits or parallel copies for data-centre setups; at home, it models layer-splitting across cards and memory spill into system RAM. Its stated speed estimates are usually within ±30%, and it advises benchmarking before purchase.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
○OpenPublic substance visible
✦PrettyNotable craft visible
●ProPolished or operationally mature
◎NicheUnusually specific use
One card from the complete issue