Have a ComfyUI workflow? Get a readable Workflow Doc in your browser.Make a workflow readable.Try it now ↗
stabilityai-ai-community
Runtime profile
Source description
screenshot
AI-generated commentary
AI-generated explanation based on source and configuration details. Suggestions are clearly labeled.
Sd3.5 Large ControlNet is a ComfyUI workflow for generating images from text with ControlNets, using a text prompt and a loaded image to produce previewable and saved image outputs.
The documented example uses Canny control; blur or depth are also named options.
The workflow includes image loading, ControlNet application, sampling, VAE decoding, previewing, and saving.
The workflow names text encoding and image loading, with Canny and ControlNet components for control-image processing.
PreviewImage and SaveImage are the listed image-output nodes.
For one-time setup, place sd3.5_large_controlnet_canny.safetensors (8.65GB) in ./ComfyUI/models/controlnet, sd3.5_large-q4_0.gguf (4.77GB) in ./ComfyUI/models/unet, and clip_g.safetensors (1.39GB) plus clip_l.safetensors (246MB) in ./ComfyUI/models/clip.
Also place t5xxl_fp8_e4m3fn.safetensors (4.89GB) in ./ComfyUI/models/clip and diffusion_pytorch_model.safetensors (168MB) in ./ComfyUI/models/vae.
The workflow lists these node names: ConditioningZeroOut, EmptySD3LatentImage, ModelSamplingSD3, ControlNetApplyAdvanced, ConditioningSetTimestepRange, KSampler, VAEDecode, ConditioningCombine, UnetLoaderGGUF, LoadImage, SaveImage, Canny, PreviewImage, CLIPTextEncode, TripleCLIP
The named model files are sd3.5_large_controlnet_canny.safetensors, sd3.5_large-q4_0.gguf, clip_g.safetensors, clip_l.safetensors, t5xxl_fp8_e4m3fn.safetensors, and diffusion_pytorch_model.safetensors.
The GGUF example is labeled recommended, and a separate workflow for the original safetensors is also provided.
The workflow is based on stabilityai/stable-diffusion-3.5-controlnets and is published by Calcuis.
The instructions describe running the .bat file in the main directory and dragging the workflow JSON into a browser.
The original safetensors option is described as very demanding, with good hardware advised for it.
Check the license before use: it is listed as “other,” with stabilityai-ai-community named and a LICENSE link provided.
Suggestion · not verified
Does this need to be edited?
Sign in to send an edit request.
Sources
1 sourceSource excerpts
2 excerptsEstimated VRAM requirement
Estimate unavailable
23.8 GB across 5 of 6 model files. Model file total + 25% loading overhead + 2 GB execution buffer, rounded up.
Requirements
23 requirementsText encoder · 1.29 GB · SAFETENSORS · Hugging Face · nekofantasia/nekofantasia-alpha
Text encoder · 235 MB · SAFETENSORS · Hugging Face · diamkan/comfyui_oldmodel
VAE · 9.61 GB · Hugging Face · Kwai-Kolors/Kolors
ControlNet · 8.06 GB · SAFETENSORS · Hugging Face · stabilityai/stable-diffusion-3.5-controlnets
sd3.5_large-q4_0.gguf
Not resolvedCheckpoint · Unknown
Text encoder · 4.56 GB · SAFETENSOR · fp8 · Hugging Face · comfyanonymous/flux_text_encoders
Canny
Not resolvedNode pack · Unknown
CLIPTextEncode
Not resolvedNode pack · Unknown
ConditioningCombine
Not resolvedNode pack · Unknown
ConditioningSetTimestepRange
--- license: other license_name: stabilityai-ai-community license_link: LICENSE language: - en base_model: - stabilityai/stable-diffusion-3.5-controlnets pipeline_tag: text-to-image tags: - stable-diffusion - controlnet - gguf-comfy --- # GGUF quantized version of Stable Diffusion 3.5 Large with ControlNets  ## Setup (once) - drag sd3.5_large_controlnet_canny.safetensors ([8.65GB](https://huggingface.co/calcuis/sd3.5-large-controlnet/blob/main/sd3.5_large_controlnet_canny.safetensors)) to > ./ComfyUI/models/controlnet - drag sd3.5_large-q4_0.gguf ([4.77GB](https://huggingface.co/calcuis/sd3.5-large-controlnet/blob/main/sd3.5_large-q4_0.gguf)) to > ./ComfyUI/models/unet - drag clip_g.safetensors ([1.39GB](https://huggingface.co/calcuis/sd3.5-large-controlnet/blob/main/clip_g.safetensors)) to > ./ComfyUI/models/clip - drag clip_l.safetensors ([246MB](https://huggingface.co/calcuis/sd3.5-large-controlnet/blob/main/clip_l.safetensors)) to > ./ComfyUI/models/clip - drag t5xxl_fp8_e4m3fn.safetensors ([4.89GB](https://huggingface.co/calcuis/sd3.5-large-controlnet/blob/main/t5xxl_fp8_e4m3fn.safetensors)) to > ./ComfyUI/models/clip - drag diffusion_pytorch_model.safetensors ([168MB](https://huggingface.co/calcuis/sd3.5-large-controlnet/blob/main/diffusion_pytorch_model.safetensors)) to > ./ComfyUI/models/vae ## Run it straight (no installation needed way) - run the .bat file in the main directory (assuming you are using the gguf-comfy pack below) - drag the workflow json file (below) to > your browser ### Workflows - example workflow for [gguf](https://huggingface.co/calcuis/sd3.5-large-controlnet/blob/main/workflow-sd3.5-controlnet-gguf.json) (recommended; demo above is canny; free to opt blur or depth) - example workflow for the original safetensors (this option is very demanding; make sure you have good hardware) ### References - base model from [stabilityai](https://huggingface.co/stabilityai/stable-diffusion-3.5-controlnets) - comfyui from [comfyanonymous](https://github.com/comfyanonymous/ComfyUI) - gguf node from [city96](https://github.com/city96/ComfyUI-GGUF) - gguf-comfy [pack](https://github.com/calcuis/gguf-comfy/releases)
Source context: 635 downloads · 9 likes · Pipeline text-to-image · Library diffusers · Repo calcuis/sd3.5-large-controlnet
Node pack · Unknown
ConditioningZeroOut
Not resolvedNode pack · Unknown
ControlNetApplyAdvanced
Not resolvedNode pack · Unknown
ControlNetLoader
Not resolvedNode pack · Unknown
EmptySD3LatentImage
Not resolvedNode pack · Unknown
KSampler
Not resolvedNode pack · Unknown
LoadImage
Not resolvedNode pack · Unknown
ModelSamplingSD3
Not resolvedNode pack · Unknown
PreviewImage
Not resolvedNode pack · Unknown
SaveImage
Not resolvedNode pack · Unknown
TripleCLIPLoader
Not resolvedNode pack · Unknown
UnetLoaderGGUF
Not resolvedNode pack · Unknown
VAEDecode
Not resolvedNode pack · Unknown
VAELoader
Not resolvedNode pack · Unknown
