Open Source Model Manager

mods

The package manager for AI models.
Install checkpoints, LoRAs, VAEs, and ControlNets with a single command.

Terminal
$ mods install z-image-turbo --variant gguf-q4-k-m
→ Install plan for z-image-turbo:
 
Z-Image Text Encoder (text_encoder) [bf16] 8.04 GiB
Z-Image VAE (vae) 335.00 MiB
Z-Image-Turbo (diffusion_model) [gguf-q4-k-m] 4.98 GiB
 
Total download: 5.31 GiB
 
Installed z-image-turbo successfully.

Why mods?

Stop manually downloading models. Stop juggling folders. Start shipping.

📦

One Command Install

Install any model with a single command. Dependencies resolve automatically.

🔗

Content-Addressed Storage

Models stored once, symlinked everywhere. No duplicates across ComfyUI, A1111, or InvokeAI.

Smart Variant Selection

Automatically picks the right fp16/fp8/GGUF variant for your GPU's VRAM.

🔄

Dependency Resolution

Install a ControlNet and its base model, text encoder, and VAE all come along.

🔍

Search & Discover

Browse 61 models and growing. Filter by type, architecture, tags, or rating.

🩺

Health Checks

mods doctor finds broken symlinks, missing deps, and corrupt files instantly.

How It Works

1

Search

$ mods search flux

Browse 61 models across checkpoints, LoRAs, VAEs, ControlNets, and more.

2

Install

$ mods install flux-dev --variant fp8

One command. Dependencies auto-resolve. Smart VRAM-based variant selection.

3

Use

$ mods list

Models are symlinked into your ComfyUI, A1111, or InvokeAI folders. Zero config needed.

Model Registry

61 models and growing. Image, video, upscalers, LoRAs — all one command away.

FLUX

State-of-the-art image generation

  • FLUX.2 Dev
  • FLUX.2 Klein 4B
  • FLUX.2 Klein 9B
  • FLUX.1 Kontext Dev
  • FLUX.1 Dev
  • FLUX.1 Schnell
  • FLUX Fill Dev
  • FLUX Canny Dev
  • FLUX Redux Dev
$ mods install flux2-dev

Qwen

Multimodal editing & generation

  • Qwen Image
  • Qwen Image Edit
  • Qwen Image Edit Lightning LoRA
$ mods install qwen-image

Wan

Video generation & editing

  • Wan 2.1 VACE 14B
  • Wan 2.2 TI2V 5B
  • Wan 2.2 T2V 14B
  • Wan 2.2 I2V 14B
$ mods install wan21-vace-14b

LTX-Video

Fast video generation

  • LTX-Video 2B
  • LTX-Video 13B
  • LTX-2 19B
$ mods install ltx-video-2b

Plus 42+ more

SDXL Base 1.0 SD 1.5 Z-Image Turbo ControlNets IP-Adapters 4x UltraSharp BiRefNet LCM LoRAs

Browse all 61 models in the model library

Get Started in 30 Seconds

Install

curl -fsSL https://raw.githubusercontent.com/modshq-org/mods/main/install.sh | sh

Or build from source:

git clone https://github.com/modshq-org/mods
cd mods && cargo install --path .

Link your tools

mods link --comfyui ~/ComfyUI

Start installing

mods install sdxl-base-1.0

Before & After

Without mods

  • Search HuggingFace/Civitai manually
  • Download 10+ files for one workflow
  • Figure out which folder each goes into
  • Duplicate files across A1111 and ComfyUI
  • Forget which VAE goes with which model
  • Run out of disk space from duplicates

With mods

  • mods search flux
  • mods install flux-dev — deps included
  • Auto-placed in the right folders
  • Symlinks share one copy across tools
  • Dependencies tracked automatically
  • mods gc reclaims unused space