ostris GitHub avatar

ComfyUI-Krea2-Ostris-Edit

ostris

ComfyUI custom nodes to run Krea 2 edit LoRAs trained with ai-toolkit, enabling image editing via reference images and model patching.

Stars

233

7-day growth

No data

Forks

5

Open issues

4

License

MIT

Last updated

2026-07-17

AI repository intelligence
FR-AI / ANALYSIS

Why it is worth attention

It brings Krea 2 edit LoRAs into ComfyUI without extra dependencies, with efficient reference encoding and optional KV cache for faster inference.

Who it is for

  • ComfyUI users
  • AI image editing enthusiasts
  • Developers using ai-toolkit for training LoRAs
  • Researchers exploring diffusion model conditioning

Use cases

  • Edit an image by providing reference images via the Text Encode node
  • Apply a Krea 2 edit LoRA in a ComfyUI workflow
  • Experiment with reference-aware conditioning and model patching
  • Speed up inference with KV caching for compatible LoRAs

Strengths

  • No extra dependencies required beyond cloning the repo
  • Patched model behaves like stock if no reference latents, safe to leave in graph
  • KV cache precomputes reference token attention for speed
  • Clear example wiring and node descriptions

Considerations

  • Text encoder checkpoint must include Qwen3-VL vision weights
  • KV cache only works if LoRA was trained with ai-toolkit's kv_cache model kwarg
  • Reference images are downscaled to fit 384x384 pixels

README quick start

Installation

Description

Comfy UI Nodes for Krea 2 LoRAs trained with AI Toolkit Experimental Edit

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