Updated: May 17, 2025
base modelVerified: 6 months ago
GGUF
The FLUX.1 [dev] Model is licensed by Black Forest Labs. Inc. under the FLUX.1 [dev] Non-Commercial License. Copyright Black Forest Labs. Inc.
IN NO EVENT SHALL BLACK FOREST LABS, INC. BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH USE OF THIS MODEL.
π FLUX.1-Fill-dev Quantized: Advanced Diffusion for Seamless Image Editing
π Overview
FLUX.1-Fill-dev Quantized represents a breakthrough in efficient diffusion models for image editing. Built on Black Forest Labs' original architecture and optimized through GGUF quantization, it delivers professional-grade inpainting and outpainting capabilities while maintaining impressive performance even on consumer hardware.
π Key Features
Optimized Performance: Multiple quantization options (Q8, Q5_K_M, Q4_K_M) to balance quality and speed
Versatile Editing: Excels at both inpainting (filling missing areas) and outpainting (extending images beyond boundaries)
Seamless Integration: Compatible with popular frameworks like ComfyUI and other diffusion workflows
Memory Efficient: Reduced model size without significant quality degradation
β‘ Available Versions
Q8 Quantization: Highest quality, larger model size (recommended for high-end GPUs)
Q5_K_M Quantization: Balanced performance and quality
Q4_K_M Quantization: Fastest performance, smallest size (ideal for lower-end hardware)
π Recommended Setup
π ComfyUI/
βββ π models/
β βββ π diffusion_models/
β β βββ π fluxfill-dev-q8.gguf (or q5km or q4km)
β βββ π text_encoders/
β β βββ π clip_l.safetensors
β β βββ π t5xxl_fp8_e4m3fn.safetensors (fp16 or fp8 scaled)
β βββ π vae/
β β βββ π ae.safetensors
π Getting Started
System Requirements
Minimum: 8GB VRAM (with Q4_K_M)
Recommended: 12GB+ VRAM (for Q8 version)
Optimal: 16GB+ VRAM (for complex workflows)
Installation Steps
Download your preferred quantization version
Place the model file in your
ComfyUI/models/diffusion_models/
directoryDownload required text encoders from Hugging Face
Download the VAE from Hugging Face
Note: You only need to choose ONE of the T5XXL options below based on your hardware capabilities
T5XXL Options (choose only one):
π Optimal Workflows
FLUX.1-Fill-dev excels at seamlessly patching areas within images while maintaining perfect context awareness:
Load your image into ComfyUI
Create or upload a mask for the area to fill
Use 20-30 steps with a sampling method like DPM++ 2M Karras
Extend your images beyond their original boundaries with natural continuity:
Load your image into ComfyUI
Use the FluxFill outpainting workflow
Choose the direction(s) to extend the image
π Credits
Special thanks to Black Forest Labs for developing the original FLUX.1-Fill-dev model.
π¨βπ» Developer Information
This workflow guide was created by Abdallah Al-Swaiti:
For additional tools and updates, check out the OllamaGemini Node: GitHub Repository
β¨ Elevate Your Creative Vision β¨