ComfyUI Models & Workflows Repository

This repository hosts a curated collection of ComfyUI workflows, Flux Kontext foundation models, GGUF quants, SDXL checkpoints, and LoRAs.

• Source Volume: comfyui (vo-JkzAbsxlXNZocsQDRIOIbY)
• Total Assets: 125 files (~119.87 GB)
• Updated at: 2026-09-17 09:19:07 UTC

---

📂 Repository Contents

1. Foundation Models & Checkpoints

File Path	Size	Description
models/diffusion_models/flux1-kontext-dev.safetensors	22.17 GB	Full precision Flux Kontext Dev
models/diffusion_models/flux1-dev-kontext_fp8_scaled.safetensors	11.09 GB	Scaled FP8 Flux Kontext Dev
models/gguf/flux1-kontext-dev-Q8_0.gguf	11.84 GB	Q8_0 GGUF Quantized
models/gguf/flux1-kontext-dev-Q6_K.gguf	9.17 GB	Q6_K GGUF Quantized
models/gguf/flux1-kontext-dev-Q5_K_M.gguf	7.84 GB	Q5_K_M GGUF Quantized
models/gguf/flux1-kontext-dev-Q4_K_M.gguf	6.46 GB	Q4_K_M GGUF Quantized
models/clip/t5xxl_fp8_e4m3fn_scaled.safetensors	4.80 GB	T5XXL Text Encoder (Scaled)
models/clip/clip_l.safetensors	234.7 MB	CLIP-L Text Encoder
models/vae/ae.safetensors	319.8 MB	Flux Autoencoder (VAE)

2. Workflows (19 Workflows)

• Found under workflows/:
  • workflows/Flux_kontext/ (Kontext 3D scene, character migration, relighting, etc.)
  • workflows/TY/ (Background + model workflows)

3. LoRAs (92 LoRA Models)

• Found under models/loras/:
  • Style LoRAs: Ghibli, Synthwave, Pixel, Paper cut, Oil painting, Retro manga.
  • Functional LoRAs: Relighting, Pose control, Deblur, High-res fix, Background replacement.

---

🚀 How to Use in ComfyUI

1. Download the required model to your local or cloud ComfyUI directory:

# Example: Download via huggingface-cli
huggingface-cli download zAura-xec/comfyui models/diffusion_models/flux1-dev-kontext_fp8_scaled.safetensors --local-dir /path/to/ComfyUI/models/diffusion_models --local-dir-use-symlinks False

2. Load any workflow JSON from workflows/ directly in ComfyUI by dragging and dropping the file into the canvas.