This is your work, valued

Generative AI researcher building practical diffusion model tools for image synthesis, control, and video generation.

Diffusion model control and conditioningPractical generative UI and promptingVideo diffusion and frame synthesisAgent conversation systems and compilationLayer and transparency in image generationSketch-to-image and artistic style transfer

inferred from the public footprint

4
Wed
Thu
Fri
Sat
Sun
Mon
Yesterday
Active 2d ago

lllyasviel

Top 3%
@lllyasviel

Zhang focuses on making diffusion models work in the real world—from streamlining prompt workflows in Fooocus to tackling video generation practicality in FramePack and relighting challenges in IC-Light. 169k GitHub stars across projects reflects sustained output in generative AI tooling.

FramePack. I made video generation work on laptops like image generation does.

17k

Fooocus. I made software that turns your descriptions into better AI images.

51k

IC-Light. I made a tool that changes how light falls on objects in photos.

8.5k

Paints Undo. I built an AI model that reverses finished paintings into their drawing steps.

4.1k

Forge. I made a faster, friendlier version of the most popular image AI tool.

13k

ControlNet. I made a tool that guides AI image generation with visual references.

18k

Omost. I built an AI that writes code to compose images from parts.

7.6k

ControlNet. I made a tool that lets you guide AI image generation with custom conditions.

34k

LuminaBrush. I built a tool that lets you draw lighting effects onto any image.

776

Layer Diffusion. I made an AI image tool that generates transparent layers instead of solid backgrounds.

4.1k

YGOPro Unity. I built a card game engine in Unity tested by 100,000 players.

346

VCC. I built a tool that organizes and searches your AI conversation logs.

392

LayerDiffuse. I made AI that generates images with transparent layers.

2.2k

ControlNet. I made AI image generation listen to your sketches and drawings.

5.2k

Style2Paints. I made software that colors your sketches using any art style you choose.

18k

Lvmin Zhang. I made AI tools for image editing and generation.

52

LayerDiffuse. I made an AI tool that separates photos into transparent layers.

422

Style2Paints. I built a research website exploring AI-powered artistic style transfer for paintings.

34

HuggingFace Guess. I made a tool that identifies which AI model file you're working with.

50

SingleFileDB. I made a database that fits in one tiny file with no setup.

85

PaintingLight. I made software that relights finished digital paintings.

738

Forge Legacy Extensions. I archived a collection of legacy Forge extensions for reference.

15

Blockly Prototypes. I built a space to test experimental features using visual blocks.

13

ToonDecompose. I built a tool that automatically splits cartoon animations into separate moving objects.

130

DanbooRegion. I created a dataset of illustrated images with artist-drawn region labels.

400

Appearance Eraser. I built an image smoothing toolkit that separates texture from shape.

186

Discussion. I built a discussion space for people collaborating on AI projects.

12

VCC Experiments. I built a system that teaches AI agents to learn from their own mistakes.

16

SketchKeras. I built a tool that turns paintings into hand-drawn sketches.

472

Lllyasviel. I made a personal website to share my projects and work.

22

Pages. I made a way to host websites for free using GitHub.

3

Misc Files. I made a place to store miscellaneous files together.

3

Split Filling. I made a tool that fills line art when you guide it with color strokes.

27

MangaFilter. I built a tool that transforms illustrations into manga artwork.

45

Toon Decompose. I made a tool that converts photos into cartoon layers.

7

MangaCraft. I made software that converts photos into manga artwork.

698