Parameters & inputs
What every input means and how it moves the result.
-
01
Model input: catalog, HuggingFace, manual
Three ways to describe a model and what the calculator reads from a HuggingFace repo.
3 min read · updated 2026-10-26 -
02
Weight and KV-cache precision
The weight dtypes per runtime, the KV-cache dtypes and where each is native.
4 min read · updated 2026-10-26 -
03
Agentic session parameters
The session profile fields of agentic mode: meaning, where to measure each and what it moves.
5 min read · updated 2026-10-26 -
04
Workload: context, users, batch, traffic
How context length, concurrent users, batch and traffic shape define the load you size for.
4 min read · updated 2026-10-26