Wan Animate 2 ComfyUI | Precise Motion Transfer
Wan-Animate-2 turns one still into powerful full-body motion - no skeletons.
AI model workflow
WAN ComfyUI Workflows
Create AI videos with ready-to-run WAN ComfyUI workflows for text-to-video, image-to-video, character animation, lip sync, camera control, and video editing. Run them on RunComfy Cloud, download workflow JSON, or deploy supported workflows as APIs.
51 workflows
Featured workflow
Wan 2.2 | Open-Source Video Gen Leader
The workflows
51 WAN ComfyUI Workflows
Each WAN ComfyUI example below is a real video workflow for Wan 2.1, Wan 2.2, Wan Animate, Wan VACE, Wan S2V, Wan FLF2V, Wan Fun, Wan Lightning, Wan FusionX, Wan Alpha, MultiTalk, InfiniteTalk, Phantom, Ditto, LoRA, ControlNet, text-to-video, image-to-video, and video-to-video editing. Run it on RunComfy Cloud, download the workflow JSON, or deploy supported workflows as APIs.
Wan Dancer | Music-to-Dance Video Generator
Turns image and music into perfectly synced dance videos fast.
Character & Pose & Background Replacement V3 | Wan2.2 Animate + SAM3.1 + SDPose
Swap characters, poses, and scenes fast with total creative control.
WAN 2.2 Smooth Workflow v5.0 | AI Video Generator
Turn text or images into silky-smooth short videos instantly.
InfiniteTalk | Lip-Synced Avatar Generator
Photo + Voice = Perfectly Synced Talking Avatar in Minutes
Wan 2.2 Animate | Character Swap & Lip-Sync
Transforms any face to speak and move like the original with ease.
Wan 2.2 FLF2V | First-Last Frame Video Generation
Generate smooth videos from a start and end frame using Wan 2.2 FLF2V.
Wan2.2 Animate Action Transfer V7 | AI Motion Video Maker
Makes your characters move exactly like real people—fast and controlled.
Wan 2.2 Lightning T2V I2V | 4-Step Ultra Fast
Wan 2.2 now 20x faster! T2V + I2V in 4 steps.
Wan 2.2 VBVR in ComfyUI | Video-Based Video Reasoning
Creates smart, consistent videos with real scene logic and structure.
Stable Video Infinity 2.0 | Long-Form Video Generator
Create long, smooth, story-driven AI videos effortlessly.
Wan 2.2 Low Vram | Kijai Wrapper
Low VRAM. No longer waiting. Kijai wrapper included.
Wan 2.2 + Lightx2v V2 | Ultra Fast I2V & T2V
Dual Light LoRA setup, 4X faster.
VACE 14B: All-in-One Video Creation & Editing
Create, edit and transform videos with the powerful VACE Wan2.1 14B.
Wan 2.2 Animate V2 | Realistic Pose Video Generator
Transforms photos into smooth-motion animated character videos using Wan 2.2.
Wan 2.1 | Revolutionary Video Generation
Create incredible videos from text or images with breakthrough AI running on everyday CPUs.
Wan Alpha | Transparent Video Generator
Alpha magic: instant transparent background videos for VFX and design.
Wan 2.2 | Open-Source Video Gen Leader
Available now! Better precision + smoother motion.
Create Coherent Scenes | Consistent Story Art Generator
Build seamless storytelling scenes with rich visual consistency.
Wan 2.1 Fun | ControlNet Video Generation
Generate videos with ControlNet-style visual passes like Depth, Canny, and OpenPose.
Pose Control LipSync S2V | Expressive Video Generator
Turn images into talking, moving characters with pose and audio control.
Wan 2.2 Qwen Multiangle Split Stack | Multi-View Video Generator
Turn one image into synced multi-angle animated videos fast.
Multitalk | Realistic Talking Video Maker
One-click create multi-speaker lip-sync videos from portraits and voices!
Wan2.2 VACE Fun | Image to Animated Video
Turn still photos into lifelike animated videos with custom prompts.
Wan2.2 S2V | Sound to Video Generator
Turns your audio clip into lifelike, synced video from one image
Instagirl v.20 | Wan 2.2 LoRA Demo
A Wan 2.2 workflow for demoing the Instagirl LoRA by Instara.
Wan 2.1 Ditto | Cinematic Video Restyle Generator
Transform videos into stunning artistic styles with perfect motion flow.
Wan 2.2 VACE | Pose-Controlled Video Generator
Turn still images into stunning motion with pose-based control.
Wan 2.2 Image Generation | 2-in-1 Workflow Pack
MoE Mix + Low-Only with upscale. Pick one.
Wan 2.2 Video Restyle | First Frame Restyle for Consistent and Cinematic Video Generation
Change the first frame, folks, your style makes the whole video look amazing. Pure magic.
Wan2.1 Stand In | Consistent Character Video Maker
Keeps characters consistent across video from just one reference image.
Wan 2.1 Fun | I2V + T2V
Empower your AI videos with Wan 2.1 Fun.
32 of 51 workflows shown
About WAN
What is a WAN ComfyUI workflow?
A WAN ComfyUI workflow is a reusable video node graph built around Alibaba's Wan model family. It connects a compatible Wan checkpoint with prompts, images, video or audio inputs, sampling and decoding nodes, plus any motion LoRA, VACE, ControlNet, interpolation, or upscaling steps the result needs. You can start from a working graph instead of wiring a video pipeline from scratch.
Choose Wan 2.2 for the newest workflows on this page, including text-to-video, image-to-video, first-and-last-frame generation, speech-to-video, and Wan Animate motion transfer. Wan 2.1 remains useful for established LoRA, ControlNet, VACE, camera-motion, and video-to-video graphs. The correct checkpoint, resolution, input type, and quantization depend on the workflow, so use a graph made for the same Wan version rather than swapping models blindly.
On RunComfy, each WAN ComfyUI workflow page shows a real output, the intended inputs, and the files the graph requires. Click Run now to open the prepared workflow in cloud ComfyUI, or download the WAN workflow JSON for your own installation. Local use still requires the matching Wan weights, text encoders, VAE, and custom nodes.
References and model resources: Wan2.2 GitHub, Wan2.1 GitHub, Explore WAN models on Hugging Face, Explore WAN models on Civitai.
Capabilities
What you can create with these WAN workflows
WAN ComfyUI workflows cover the full video path: prompt and image generation, controllable motion, audio-driven characters, keyframe guidance, and video editing.
Text-to-video and image-to-video
Generate a video from a written prompt or animate a starting image with a general WAN 2.2 ComfyUI workflow.
Character animation and motion transfer
Transfer body motion from a reference video to a character image while preserving appearance and scene identity.
Speech-to-video and lip sync
Drive a portrait or character video from speech, sound, or dialogue with synchronized facial movement.
Keyframe and structure control
Control timing and composition with first and last frames, pose guidance, depth, edges, or motion trajectories.
Video editing and restyling
Edit, extend, repaint, or restyle existing footage with VACE and video-to-video WAN workflows.
WAN models on RunComfy
Choose the right WAN video model. Run the complete graph.
RunComfy opens WAN ComfyUI workflows with their video model, encoders, VAE, LoRAs, control models, and custom nodes prepared. Upload the prompt, image, video, pose, or audio input the graph expects, then change motion, duration, resolution, guidance, or camera settings in native ComfyUI. Use a cloud GPU for heavier Wan 2.2 and long-video pipelines instead of rebuilding a reduced local setup.
Run a WAN workflow nowNative ComfyUI graphs for WAN video generation
Open Wan 2.1 and Wan 2.2 text-to-video, image-to-video, FLF2V, VACE, Animate, and speech-to-video pipelines as editable node graphs.
WAN checkpoints, LoRAs, and control models in the cloud
Keep compatible Wan weights, motion LoRAs, VACE and ControlNet models, and supporting files available with the cloud workflow.
Custom nodes for motion, audio, and long video
Use the nodes required for lip sync, pose transfer, camera control, interpolation, restyling, and long-form continuation without reconstructing the environment.
Cloud GPUs for larger WAN video pipelines
Match GPU memory to the chosen Wan model, frame count, resolution, and control stack instead of cutting the workflow down to fit local hardware.
Applications
What people use these WAN workflows for
Choose by the video you want to publish, from animated portraits and talking avatars to dance clips, controlled camera shots, story scenes, and restyled footage.
Photo Animation
Turn a still portrait or character image into a natural moving video while keeping the subject recognizable.
Talking Avatars
Create a lip-synced presenter, virtual character, or dialogue clip from a portrait and an audio track.
AI Dance Videos
Translate music and choreography into full-body character motion for dance and performance clips.
Cinematic Camera Moves
Add controlled pans, zooms, rotations, and push-in shots to a source image for cinematic motion.
Consistent Story Scenes
Build connected shots that keep characters, visual style, and scene details coherent across a sequence.
Video Restyling
Transform footage into a new cinematic or illustrated look while retaining the original motion.
Frequently asked questions
WAN workflow FAQ
Answers for choosing Wan 2.2 or Wan 2.1 workflows, using S2V lip sync, Wan Animate, VACE video editing, character consistency, VRAM, and JSON downloads.
What WAN ComfyUI workflow should I start with?
Start with a Wan 2.2 image-to-video or text-to-video workflow for general video generation. Use a Wan Animate workflow when you want character replacement or motion transfer, a Wan S2V workflow when speech should drive the performance, and a VACE workflow when you need video editing or restyling. Match the workflow to the input you already have.
Should I use Wan 2.2 or Wan 2.1?
Use Wan 2.2 for newer text-to-video, image-to-video, first-last-frame, Animate, and speech-to-video workflows. Use Wan 2.1 when the workflow depends on older VACE, ControlNet, LoRA, camera-control, or low-VRAM community setups. The version matters because the checkpoint, VAE, scheduler, and custom nodes can differ.
Can Wan 2.2 make lip-sync or speech-to-video?
Yes. Wan 2.2 S2V workflows use an image and audio input to create an audio-driven video, and some workflows add pose or character replacement control. Use a dedicated S2V or lip-sync graph instead of a basic image-to-video graph when mouth movement needs to follow speech or singing.
What is Wan Animate used for?
Wan Animate is used for character animation and replacement: you provide a character image and a driving video, then the workflow transfers body motion, expression, and scene movement. It is a better fit than plain image-to-video when the goal is to replace a performer or keep a character following a specific reference motion.
What is VACE in a WAN ComfyUI workflow?
VACE is the WAN video editing path for changing existing footage, such as restyling, repainting, extending, or guiding video with masks and control inputs. Use it when you already have a clip and want controlled edits, not when you only need a fresh text-to-video generation.
How do I keep a character consistent in WAN videos?
Use a reference image workflow, Wan Animate, pose-guided control, or a compatible Wan LoRA when the same person or character must stay consistent. For longer stories, generate shorter connected clips and carry forward a reference frame, because identity can drift across long generations. Train a WAN LoRA when one character needs to appear across many videos.
How much VRAM do I need for Wan 2.2 in ComfyUI?
Wan 2.2 VRAM needs depend on model size, resolution, frame count, quantization, and whether the graph adds Animate, S2V, VACE, ControlNet, or upscaling. Low-VRAM workflows usually reduce resolution or use quantized weights. Use Run now if you want the complete prepared graph on a cloud GPU, or Download JSON if you want to run it locally with the matching model files.
Related resources
More ways to use WAN
Train a WAN LoRA with a supported Wan 2.1 or Wan 2.2 base model, or open a matching RunComfy Models page for direct generation.
Train a Wan 2.1 1.3B LoRA
Use the smaller Wan 2.1 text-to-video base for a lighter custom video LoRA training path.
Open Wan 2.1 1.3B TrainerTrain a Wan 2.1 14B LoRA
Train a higher-capacity Wan 2.1 text-to-video LoRA for reusable motion or visual concepts.
Open Wan 2.1 14B TrainerTrain a Wan 2.1 I2V 480P LoRA
Train against the Wan 2.1 14B image-to-video 480P base when a source image drives the clip.
Open I2V 480P TrainerTrain a Wan 2.1 I2V 720P LoRA
Use the Wan 2.1 14B image-to-video 720P base for higher-resolution image-driven training.
Open I2V 720P TrainerTrain a Wan 2.2 14B T2V LoRA
Train a custom text-to-video LoRA on the Wan 2.2 14B base.
Open Wan 2.2 T2V TrainerTrain a Wan 2.2 14B I2V LoRA
Train a Wan 2.2 image-to-video LoRA for concepts driven by a reference image.
Open Wan 2.2 I2V TrainerTrain a Wan 2.2 TI2V 5B LoRA
Use the unified 5B base for a smaller Wan 2.2 text-and-image-to-video training path.
Open Wan 2.2 5B TrainerUse WAN models directly
Open the WAN Models page when you want direct generation instead of a complete workflow graph.
Browse WAN ModelsExplore Wan 2.2 models
Compare models and generation options grouped specifically around Wan 2.2.
Browse Wan 2.2 ModelsExplore Wan 2.5 models
Open the Wan 2.5 Models page for direct access to the newer model line where available.
Browse Wan 2.5 Models