-webkit-font-smoothing: antialiased; -moz-osx-font-smoothing: grayscale; font-smooth: never;

Welcome to local, personal AI

LMNC™ v4.20 Available

Run models up to

100

Billion Params

Monthly energy costs to run local inference

Truffle–1

$8.28

Apple M1

$19.87

RTX 4090

$74.52

64

GB of RAM

Tokens/s on Mixtral8x7B

M1 Mac • 8

RTX 3090 • 18

Truffle–1 • 20

Run multiple models at a time

Playground

Experiment with models directly in-app

Model Training Center

Discover, Install, Run.

All in a few clicks.

+6k

Models Available

Works on

Cuda

Metal

Vulcan

Search & Explore

Find the best models for particular task