Starnodes2024 GitHub avatar

comfyui-starnodes-modelconverter

Starnodes2024

A ComfyUI custom node suite for converting, quantizing, analyzing, and managing diffusion models with profile-based mixed-precision quantization and advanced layer analysis.

Stars

68

7-day growth

No data

Forks

5

Open issues

3

License

MIT

Last updated

2026-07-14

AI repository intelligence
FR-AI / ANALYSIS

Why it is worth attention

It introduces profile-based quantization that extracts blueprints from optimized models, offers deep quantization via comfy-kitchen, and includes dedicated nodes for splitting AIO checkpoints and detailed layer analysis, all within ComfyUI.

Who it is for

  • ComfyUI power users who need model optimization
  • AI artists working with large diffusion models
  • Machine learning engineers exploring model compression
  • Developers building workflows with mixed-precision formats

Use cases

  • Compress a 20GB+ Flux model to 4-bit (NVFP4) for faster inference
  • Split an all-in-one checkpoint into separate diffusion, text encoder, and VAE files
  • Analyze layer formats and generate a quantization profile for reuse on other models
  • Apply a saved profile to convert another model to INT4 ConvRot while preserving critical layers

Strengths

  • Supports 8 target formats including NVFP4, FP8, INT4 ConvRot, and MXFP8
  • Smart layer blacklisting preserves architecture-specific critical layers automatically
  • Profile system allows saving and applying quantization strategies across models
  • Detailed layer analysis with Normal and Tree views and profile export

Considerations

  • Requires 64GB+ RAM for large models like Flux or LTX
  • NVFP4 and certain formats need NVIDIA GPU; some require specific compute capability
  • Depends on external library comfy-kitchen for deep quantization, which adds setup complexity

README quick start

Installation

Description

Ultimate Model Converter for ComfyUI using comfyui-kitchen - Convert between Transformers, FP32, FP16, FP8. INT8, NVFP4, INT8 Comvrot

Related repositories

Similar projects matched by category, topics, and programming language.

lopopolo
Featured
lopopolo GitHub avatar

harness-engineering

Harness Engineering is a methodology for improving coding agent outputs by carefully crafting the environment around them—providing curated context, tools, and executable constraints that encode an organization’s nonfunctional requirements and cumulative lessons.

AI & Machine LearningAI Agents
2,390
slvDev
Featured
slvDev GitHub avatar

esp32-ai

A 28.9 million parameter language model runs on an $8 ESP32-S3 microcontroller entirely on-device, generating simple stories at about 9.5 tokens per second.

AI & Machine LearningLarge Language Models
1,960
littledivy
Featured
littledivy GitHub avatar

mimic

mimic captures traffic from any iOS or web app and automatically generates a Python client library that lets you call the app's API like a regular library.

AI & Machine Learning
1,482