Offline models running on local hardware. Air-gapped, no external API calls.
local-models
llama-3.3-70b
mistral-7b