Explore Community Workflows
Discover amazing workflows created by the ComfyUI community. Download, use, and get inspired by creative workflows from artists and developers around the world.
First & Last Frame i2v WAN2.2 (Native)
A basic image-to-video workflow where you provide the first and last frames as reference images, and the model will fill in the in-between animation. #i2v #flf #wan22
Text to Video WAN2.2 (Native)
A basic Text-to-Video workflow that generates a video based on your text prompt using WAN2.2 and the fast lightning LoRAs in just 4 to 7 steps. #t2v #wan22
Qwen-Image-Edit (Native)
Say goodbye to tedious manual edits! With Qwen’s powerful AI image editing workflow, you can transform images effortlessly using simple text instructions #i2i #edit
Qwen-Image Outpaint (Native)
Go beyond the edges: AI generates coherent extensions of your image based on context. No cropping needed—grow your scene. #i2i #outpainting #qwenimage
Qwen-Image Inpainting (Native)
Take control of your image: mask any area, and let AI reconstruct it exactly how you want. No Photoshop skills needed. #inpainting #i2i #qwenimage
Flux LoRAs (Native)
Flux LoRAs are small but mighty models that inject custom styles or characters into any model. LoRAs enable the customization of a model to generate new concepts or styles. #t2i #LoRA #Flux
Flux Text-To-Image (Native)
The simplest way to start with AI image generation: type a prompt, watch art emerge, powered by probabilistic sampling and smart denoising. #t2i #Flux
Flux Depth ControlNet (Native)
Depth maps are beneficial for separating elements based on camera proximity, allowing for a higher degree of change while maintaining structure. #i2i #depth #controlnet
Flux Open-Pose ControlNet (Native)
Flux Open-Pose enables the extraction of a pose from a reference image, allowing for its transfer to the generated image that follows the text prompt. #i2i #openpose #controlnet.
Flux Canny ControlNet (Native)
The Canny Maps ControlNets help replicate the underlying structure and composition of images, allowing you to stylise them easily. #i2i #canny #contrlnet
Flux Redux (Native)
Flux Redux lets you use an Image as a reference to extract its characteristics, like style, and use it to generate another image influenced by your text prompt #fluxredux #i2i #style
Flux Outpaint (Native)
This workflow provides a solution to one of the most common problems when editing images, extending the frame of an image with Flux. #i2i #outpaint #fluxFill
Flux Fill Inpaint (Native)
Basic Flux Fill inpainting workflow, with this fundamental technique, you can edit and modify elements on the image by drawing masks and typing prompts. #i2i #inpainting #fluxfill
Text to Image with Chroma (Native)
Text-To-Image workflow using Chroma. Chroma is a popular model due to its range of style variations. #t2i #chroma
Sound-To-Video WAN2.2 (Native)
A Sound-to-Video workflow uploads an audio track and an image reference; the model will create the animation and lipsync with WAN2.1 or WAN2.2 #s2v #lipsync #i2v
Qwen-Image InstantX controlnet (Native)
An Image-to-Image workflow that lets you use a reference image as the structural guidance for the composition using QwenImage and InstantX ControlNet (native) #i2i #QwenImage #controlnet
Branding Product shot with (SeedDream-v4 Dual-Input)
An image-to-image workflow based on the latest Seedream-v4. Upload an image of your product with your logo, and the model will overlay your branding label or logo on the product.
Qwen-Image-Edit Try-On
A basic Try-on and image-to-image workflow using Qwen-Image-Edit, fast using distilled lora #i2i
WAN2.2 Image-to-Video with Radial-Attention
Generate videos from an image using WAN2.2. Take advantage of Radial Attention Speed-Ups to generate videos over 20 seconds in length in almost half the time.
Basic Text-to-Image with the HunyuanImage-2.1 Model
Text-to-image generation with Hunyuan Image 2.1. This model offers superior native resolution and text adherence thanks to its glyph encoder and integrated MLLM with Dual stream DiT architecture #t2i #hunyuanimage