How shadowcz007/comfyui-mixlab-nodes works
Mixlab is a ComfyUI custom-node extension for building creative workflows and publishing selected workflows as web apps. The principal path is a user configuring a ComfyUI graph, running image, prompt, media, or 3D nodes, and viewing or saving results; the web-app mode exposes configured workflow inputs and outputs in a browser. The graph includes documented capabilities whose detailed implementations were not sampled. Edges are limited to README statements and source-index references; unsampled integrations are left unconnected where their wiring is uncertain.
Main entry points: __init__.py Node Registration · generate.py Speech Generation · semantic.py Semantic Dataset · FishSpeech.py Fish Speech Node
Components 26
Workflow Extension
- ComfyUI Runtime
- Node Registration
__init__.py - Workflow Web App
web/javascript/app_mixlab.js
Image And Prompt
- Image Operations
nodes/ImageNode.py - Background Removal
nodes/RembgNode.py - Layer Compositing
nodes/ImageNode.py - Prompt Tools
nodes/PromptNode.py - Visual Styling
nodes/Style.py - Image Interrogation
nodes/ClipInterrogator.py
Media And 3D
- Video Nodes
nodes/Video.py - Fal Video Generation
nodes/FalVideo.py - Screen Capture
nodes/ScreenShareNode.py - Image To 3D
nodes/TripoSR.py - 3D Reconstruction
nodes/tsr/system.py
Speech And Language
- LLM Chat
nodes/ChatGPT.py - Text Generation
nodes/TextGenerateNode.py - Audio Nodes
nodes/Audio.py - Fish Speech Node
nodes/FishSpeech.py - Semantic Dataset
nodes/fish_speech/datasets/semantic.py - Text Semantic Model
nodes/fish_speech/models/text2semantic/llama.py - Speech Generation
nodes/fish_speech/tools/llama/generate.py - Text Cleaning
nodes/fish_speech/clean.py
Browser Experience
- ComfyUI Frontend
web/javascript/ui_mixlab.js - Node UI Extensions
web/javascript/main_mixlab.js - 3D Viewer
webApp/lib/threeVisualizer.js
Other components
- Workflow Author
Connections 26
- Workflow Author → ComfyUI Runtime: authors workflows
README.md - ComfyUI Runtime → Node Registration: loads extension
README.md - Node Registration → Image Operations: registers nodes
__init__.py - Node Registration → Background Removal: registers nodes
__init__.py - Node Registration → Prompt Tools: registers nodes
__init__.py - Node Registration → Visual Styling: registers nodes
__init__.py - Node Registration → LLM Chat: registers nodes
__init__.py - Node Registration → Video Nodes: registers nodes
__init__.py - Node Registration → Screen Capture: registers nodes
__init__.py - Node Registration → Image To 3D: registers nodes
__init__.py - Node Registration → Audio Nodes: registers nodes
__init__.py - Node Registration → Image Interrogation: registers nodes
__init__.py - Node Registration → Text Generation: registers nodes
__init__.py - Node Registration → Fish Speech Node: registers nodes
__init__.py - Node Registration → Fal Video Generation: registers nodes
__init__.py - Node Registration → ComfyUI Frontend: adds frontend
- ComfyUI Frontend → Node UI Extensions: loads extensions
web/javascript/ui_mixlab.js - Node UI Extensions → Workflow Web App: supports app mode
- Workflow Author → Workflow Web App: uses published workflow
README.md - Fish Speech Node → Text Semantic Model: uses speech model
- Speech Generation → Text Semantic Model: loads model
nodes/fish_speech/tools/llama/generate.py - Speech Generation → Text Cleaning: cleans text
nodes/fish_speech/tools/llama/generate.py - Semantic Dataset → Text Cleaning: normalizes text
nodes/fish_speech/datasets/semantic.py - Semantic Dataset → Speech Generation: evaluated by
nodes/fish_speech/tools/llama/eval_in_context.py - Node Registration → 3D Reconstruction: registers 3D nodes
- 3D Viewer → Image To 3D: displays 3D results