Posted by admin

00:00 Intro 00:43 Example 1 01:47 Example 2 02:07 Example 3 02:27 Example 4 02:39 Example 5 02:56 Example 6 03:40 Example 7/8/9 04:10 Example 10 04:32 Example 11 04:56 Example 12 /13 05:10 First Error Example 14 05:30 Installation 07:27 Outro Workflow https://openart.ai/workflows/-/-/jcO2yXHERsFl9oCKYJ3L Checkpoints - Goes in ComfyUI/models/unet (not checkpoints) fp16 version for 16gb vram GPUs https://huggingface.co/black-forest-labs/FLUX.1-dev/blob/main/flux1-dev.sft fp 8 version for less than 16gb vram GPUs https://huggingface.co/Kijai/flux-fp8/blob/main/flux1-dev-fp8.safetensors Text encoders goes in ComfyUI/models/clip: https://huggingface.co/comfyanonymous/flux_text_encoders/tree/main VAE (ae.sft) goes in ComfyUI/models/vae: https://huggingface.co/black-forest-labs/FLUX.1-schnell/blob/main/ae.sft Download the fp8 t5xxl for degraded quality but less RAM use Launch ComfyUI with "--lowvram" arg (in the .bat file) to offload text encoder to CPU. The Flux AI model (https://github.com/black-forest-labs/flux?tab=readme-ov-file) by Black Forest Labs is a cutting-edge text-to-image model, boasting an impressive 12 billion parameters. Developed by the original team behind Stable Diffusion, Flux is designed to push the boundaries of creativity and performance in generative AI. Key Features: Enhanced Image Quality: Flux generates stunning visuals at higher resolutions Advanced Human Anatomy and Photorealism: It achieves highly realistic and anatomically accurate images Improved Prompt Adherence: The model delivers more accurate and relevant images based on user inputs Exceptional Speed: The “Schnell” variant operates up to 10 times faster, ideal for high-demand applications. Variants: FLUX.1 [pro]: The base model, available via API FLUX.1 [dev]: A guidance-distilled variant FLUX.1 [schnell]: A guidance and step-distilled variant, optimized for speed Usage: Flux can be used for both text-to-image and image-to-image generation. It is available through various platforms like Replicate and FAL. For local installation, you can clone the repository from GitHub and follow the setup instructions provided