l/rllmrun Install

LOCAL · KAGGLE · RUNPOD

Run any open model.
Anywhere it fits.

llmrun finds models, exposes every quant, matches live GPU capacity and price, and keeps paid sessions inside a hard deadline.

Build a commandView on GitHub
$gh repo clone AnassKartit/llmrun && pipx install ./llmrun

COMMAND BUILDER

Describe the run. The CLI does the rest.

The browser never rents a GPU. Copy the command, inspect live prices in your terminal, then approve.

Model type
Run on

ONE COMMAND SURFACE

Text, images, tools, and budgets.

01

Browse models

llmrun search qwen

Search Hugging Face, compare fit, then inspect available quants.

02

Run text

llmrun text MODEL

Use local hardware by default, Kaggle for free VRAM, or RunPod for live paid capacity.

03

Run Diffusers

llmrun image MODEL

Serve text-to-image or image-to-image through the same local or remote API shape.

04

Open your tool

llmrun launch webui

Configure Open WebUI, OpenCode, and other OpenAI-compatible clients automatically.