SD3.5 Large Blur

Generate images guided by blurred reference images using SD 3.5.

Download Workflow Try in Cloud API documentation

About

The SD3.5 Large Blur workflow is designed to generate high-quality images by leveraging blurred reference images as a guide. This technique utilizes the Stable Diffusion 3.5 (SD3.5) model, which is known for its ability to create detailed and visually appealing images. The workflow incorporates several key nodes, including the KSampler, CheckpointLoaderSimple, and ControlNetLoader, to efficiently manage the image generation process.

By using a blurred image as a reference, the workflow can focus on capturing the essence of the image while allowing for creative interpretation in the finer details. This approach is particularly useful for artists and designers looking to create unique and imaginative visuals based on existing images.

The workflow begins by loading the necessary models and checkpoints, such as the 'sd3.5_large_fp8_scaled.safetensors' and the 'sd3.5_large_controlnet_blur.safetensors'. These models are essential for the ControlNetApplyAdvanced node, which applies advanced control techniques to the image generation process. Users are encouraged to preprocess their images using the 'comfyui_controlnet_aux' tool to ensure optimal results. The CLIPTextEncode node allows for the integration of text prompts that guide the image generation, enabling users to specify desired characteristics, such as color and style.

The final output is decoded using the VAEDecode node and saved with the SaveImage node, providing users with a ready-to-use image that meets their creative needs.

What you provide

Reading the workflow

Runs in ComfyUI. This workflow can't run on this page. Open it in Comfy Cloud with everything loaded, or download it for your own ComfyUI.

Try in Cloud

Output

Use this workflow from code

Generated from the workflow graph and inputs when possible; stored examples are shown otherwise. Review credentials and endpoints before running code, which may incur charges.

Get API key

What you need

Runs on: Comfy Cloud
Key: COMFY_API_KEY

Needs a paid Cloud plan and credits.

Download API JSON

# Python 3.10+: pip install comfy-sdk==0.4.0
import os

from comfy_sdk import Comfy

api_key = os.environ.get("COMFY_API_KEY", "").strip()
if not api_key: raise RuntimeError("Set COMFY_API_KEY")
os.environ.setdefault("COMFY_BASE_URL", "https://cloud.comfy.org")
client = Comfy(api_key=api_key)
workflow = client.workflows.from_json({
    "3": {
        "class_type": "KSampler",
        "inputs": {
            "cfg": 4,
            "denoise": 1,
            "latent_image": [
                "33",
                0
            ],
            "model": [
                "4",
                0
            ],
            "negative": [
                "58",
                1
            ],
            "positive": [
                "58",
                0
            ],
            "sampler_name": "euler",
            "scheduler": "simple",
            "seed": 268264726798396,
            "steps": 30
        },
        "_meta": {
            "title": "KSampler"
        }
    },
    "4": {
        "class_type": "CheckpointLoaderSimple",
        "inputs": {
            "ckpt_name": "sd3.5_large_fp8_scaled.safetensors"
        },
        "_meta": {
            "title": "Load Checkpoint"
        }
    },
    "6": {
        "class_type": "CLIPTextEncode",
        "inputs": {
            "clip": [
                "4",
                1
            ],
            "text": "crystal butterfly above the sea, white, hyper detailed, with diamonds"
        },
        "_meta": {
            "title": "CLIP Text Encode (Prompt)"
        }
    },
    "8": {
        "class_type": "VAEDecode",
        "inputs": {
            "samples": [
                "3",
                0
            ],
            "vae": [
                "4",
                2
            ]
        },
        "_meta": {
            "title": "VAE Decode"
        }
    },
    "9": {
        "class_type": "SaveImage",
        "inputs": {
            "filename_prefix": "ComfyUI",
            "images": [
                "8",
                0
            ]
        },
        "_meta": {
            "title": "Save Image"
        }
    },
    "33": {
        "class_type": "EmptySD3LatentImage",
        "inputs": {
            "batch_size": 1,
            "height": 1024,
            "width": 1024
        },
        "_meta": {
            "title": "EmptySD3LatentImage"
        }
    },
    "46": {
        "class_type": "ControlNetLoader",
        "inputs": {
            "control_net_name": "sd3.5_large_controlnet_blur.safetensors"
        },
        "_meta": {
            "title": "Load ControlNet Model"
        }
    },
    "50": {
        "class_type": "ConditioningZeroOut",
        "inputs": {
            "conditioning": [
                "6",
                0
            ]
        },
        "_meta": {
            "title": "Conditioning Zero Out"
        }
    },
    "57": {
        "class_type": "LoadImage",
        "inputs": {
            "image": "sd3.5_large_blur_input_image.png"
        },
        "_meta": {
            "title": "Load Image"
        }
    },
    "58": {
        "class_type": "ControlNetApplyAdvanced",
        "inputs": {
            "control_net": [
                "46",
                0
            ],
            "end_percent": 1,
            "image": [
                "57",
                0
            ],
            "negative": [
                "50",
                0
            ],
            "positive": [
                "6",
                0
            ],
            "start_percent": 0,
            "strength": 1,
            "vae": [
                "4",
                2
            ]
        },
        "_meta": {
            "title": "Apply ControlNet"
        }
    }
})

# run() also passes api_key to the partner nodes in this workflow.
job = client.run(workflow, api_key=api_key)

API documentation

SD3.5 Large Blur

About

What you provide

Inside the workflow

Requires custom nodes

Use this workflow from code

What you need

Questions & answers

About

What you provide

Inside the workflow

Requires custom nodes

Use this workflow from code

What you need

Questions & answers

Do I need to preprocess my images before using this workflow?

Can I use any text prompt with this workflow?

What image formats are supported for the reference images?

How can I improve the quality of the generated images?