Click for Full guide!

Anima Made Easy: Complete ComfyUI Workflow with Wan Video

One workflow for everything you do with Anima. Roll fast drafts, hi-res the keepers, fix faces and hands, upscale, inpaint, restyle a picture, paint up to four characters with regional prompting, and turn any picture into a 5-second video with Wan 2.2. Every part sits in its own colour-coded group with a small on/off switch, a jump key, and a plain note that tells you what to click.

This version has draft-then-hires, a regional prompting engine, Wan 2.2 video, and a easier layout. Can potentially add more if needed but pretty happy with the new V2.0

Why you will like it

• Draft then hi-res. Roll small drafts in seconds. When one is a keeper, drag its PNG back in and hi-res it. The seed is shared, so the final is the same picture, only sharper.

• Jump keys Press 1 to 9, 0, minus, and equals to fly to any group. No scrolling around a giant canvas.

• Switches everywhere. A top panel lists every group. There is also a switch next to each group, small draft, hires and Adetailer boxes under the prompt, and an all-groups panel next to regional prompting.

• Notes right next to what they explain. Short and plain.

• Regional prompting that just works. Red, green, blue, and yellow masks, with ready-made mask pictures included for 2, 3, and 4 characters.

• Video included. Text-to-video and image-to-video with Wan 2.2 14B, using 4-step LoRAs so a clip takes about 4 steps.

Quick start (5 steps)

1. Install the custom nodes listed below (ComfyUI-Manager makes it one click each).

2. Put the model files in their folders (list below).

3. Drag the workflow JSON onto ComfyUI. Pick your files in the model boxes.

4. Press 2, check that A1 and A2 are on, type a prompt in the positive box and press Run.

5. Like a draft? Drag its PNG onto ComfyUI, set the DraftSampler to fixed, switch B on and press Run.

Requirements

Custom nodes (install with ComfyUI-Manager):

• ComfyUI-Impact-Pack and ComfyUI-Impact-Subpack (ltdrdata)

• rgthree-comfy (rgthree)

• ComfyUI-Lora-Manager (willmiao)

• ComfyUI_UltimateSDUpscale (ssitu)

• ComfyUI-ppm (pamparamm), for regional prompting

• ComfyUI-GGUF (city96), for the video groups only

Models:

What Folder Anima model official: https://huggingface.co/circlestone-labs/Anima or https://civitai.com/models/2458426/anima-official models/diffusion_models qwen_3_06b_base.safetensors models/text_encoders https://civitai.com/models/2400206/qwen-3-06b-baseanima

qwen_image_vae.safetensors models/vae https://civitai.com/models/1912333/qwen-image-vae

face_yolov9c.pt https://huggingface.co/Bingsu/adetailer/blob/main/face_yolov9c.pt models/ultralytics/bbox

hand_yolov9c.pt https://huggingface.co/Bingsu/adetailer/blob/main/hand_yolov9c.pt models/ultralytics/bbox

4x_foolhardy_Remacri.pth (any 2x or 4x upscale model works) models/upscale_models

Video models (optional): Wan 2.2 14B GGUF high and low noise, for both T2V and I2V (QuantStack on Hugging Face), the 4-step lightx2v LoRAs (lightx2v/Wan2.2-Distill-Loras), wan_2.1_vae.safetensors and umt5_xxl_fp8_e4m3fn_scaled.safetensors. Download the single files, not the huge original Wan-AI repos. The full table is inside the workflow.

https://huggingface.co/bullerwins/Wan2.2-I2V-A14B-GGUF/tree/main
https://huggingface.co/QuantStack/Wan2.2-T2V-A14B-GGUF/tree/main
https://huggingface.co/Comfy-Org/Wan_2.1_ComfyUI_repackaged/blob/main/split_files/text_encoders/umt5_xxl_fp8_e4m3fn_scaled.safetensors
https://huggingface.co/Comfy-Org/Wan_2.1_ComfyUI_repackaged/blob/main/split_files/vae/wan_2.1_vae.safetensors

Prompting Anima

Tag order that works best:

[quality / meta / year / safety tags] [1girl / 1boy / 1other] [character] [series] [@artist] [general tags]

Settings: 30 to 50 steps (35 is the default here), CFG 3 to 5, about 1 megapixel, sampler er_sde, euler_a or dpmpp_2m_sde. Anima's own page has more.

Regional prompting masks included

Five mask pictures are in the download. The 3584x1536 ones are for large canvases and the 1792x768 ones match the wide draft size. The workflow resizes them automatically:

• 2 characters: red on the left, green on the right.

• 3 characters: red, green, blue from left to right.

• 4 characters: red, green, blue, yellow from left to right.

The 3 and 4 character masks also come in 1792x768. The workflow uses the 3-character 1792x768 one by default.

Describe each person in their own colour box, and keep the base prompt for style and the shared scene only. Regional prompt images are included in the download so you don't have to make your own. They are the ones I use, so if you don't want the black spaces, you'll just need to throw it into Gemini or something and have it fix it, or do so yourself, of course.

Civitai metadata

This workflow saves pictures with the LoRA Manager save node, so your model and LoRAs show up in the Resources box when you post.

Tips

• Keep A1 on all the time. Everything uses it.

• Draft and final sizes come in pairs so the proportions match. The workflow has a chart for square, wide, and portrait.

• Video clips: 81 frames at 16 fps is about 5 seconds. Raise the frames, not the fps, for longer.

UPDATE v3.0: Brain Edition

The download now has TWO workflows. Same install, use whichever you like.

• Guided (Anima Made Easy v2.0): notes next to every box, jump keys, everything explained. Best for learning it.

• Brain Edition (new): no notes in the way. One central prompt with every control you need around it. Built for running a lot of pictures.

What's new in the Brain Edition

• The Brain (jump key 2). The global prompt sits in the middle. Around it: draft and hi-res buttons for wallpaper, square and phone, the draft seed (with the normal fixed / randomize dropdown), steps and CFG boxes, image to image, Adetailer, batch buttons and index boxes, and the final-upscale percentage for each shape.

• Steps and CFG boxes. Turbo-style checkpoints need about 12 steps, not 35. Change it right next to the model.

• Prompt file loader (overnight batch). Put your prompts in a text file, split them with a line of =====, turn it on and run a batch. Set repeats to get several pictures per prompt. You can add prompts to the bottom of the file while it runs.

• Batch Final Upscale (group J). Reads your batch hi-res folder, enlarges every picture and sizes it by shape using three percentage boxes (phone, wallpaper, square). Saves to its own folder and keeps the prompt and LoRA info inside the file.

• Regional prompting in the brain. All four colour prompt boxes, the sizes and the mask pictures in one place. Separate on/off for wide and tall (phone) regional pictures, plus a button to pick 2, 3 or 4 regions. No more swapping mask pictures by hand.

• First and Last Frame to Video (V3). Load a first and a last picture, describe the motion between them.

• Command to reset all batch index boxes to 0 (command palette: "Anima: set the batch index boxes to 0").

• New guide section for all of the above: see the full guide (section 13).

Install

1. Copy the whole anima_batch_loader folder into ComfyUI/custom_nodes/ and restart. The guided workflow only needs it for Batch Upgrade and the keys. The Brain Edition needs it for the prompt file loader and the batch groups.

2. Copy everything in the input folder of the download into ComfyUI/input/. It has all the regional masks (2, 3 and 4 regions, wide, tall and large), an example picture and an example prompts.txt.

3. Drag either workflow JSON onto ComfyUI.

Good to know

• Batch groups only read PNGs made by this workflow, because each prompt is read from the picture.

• Set the index boxes to 0 before every batch.

• Keys: K sends a picture to keepers.

• Found something broken? Tell me in the comments.

Small edits to the rest of the post

• Line 4 ("pretty happy with the new V2.0"): change to "pretty happy with the new V3.0".

• "Five mask pictures are in the download": there are now nine (2, 3 and 4 regions, in wide, tall and large sizes). Add: "The tall ones are for phone pictures."

• Quick start step 4 still works for the guided version. For the Brain Edition: press 2, type a prompt, press Run.

Feedback

Found a bug or have an idea? Tell me in the comments. If you make something cool with it, post it; I would love to see it.

Huge shoutout to Legendaer; without their workflow, I never would have done this. I used theirs for a long time, but it was just missing things I wanted, and it was so compact that things weren't where I wanted them. So I decided to work on my own!

v1.1 update

• Wide and square drafts in one workflow. There are now two draft groups, A2 Wide and A2 Square. Each has its own hi-res pass (B Wide and B Square). Wide is 1120x480 draft to 1792x768 final, and square is 768x768 draft to 1248x1248 final. Switch on the pair you want and leave the other pair off. You no longer have to swap workflows to change shape. The other sizes in the size chart still work: change the numbers in the Draft Size and Final Size boxes of the pair you use.

• One seed box for both drafts. It sits next to the prompt. -1 Gives a new random seed every run. Type a number to fix one, or click Use last queued seed on a draft you like. You don’t have to open the draft group to lock a seed.

• Switches next to the prompt. The draft and hi-res switches are now beside the prompt box, so you can pick a pair and enable hi-res without leaving the prompt area.

• Cleaner file names. Wide and square files are named separately, so square pictures never overwrite wide ones.

## UPDATE v2.0: portrait, batch upgrade, and Send to keepers

This is a big one. Everything is in the v2.0 download.

What's new

- Portrait pair. A2 and B now come in Wide, Square and Portrait (default 896x1152 up to 1344x1728).

- Sizes in one place. Under the prompt there is a chart of ready-made size pairs and a row of boxes per shape. Type the draft and final size you want and every group follows.

- One small box per shape, next to the prompt. Draft on top, hires below. Adetailer has its own box, so hires no longer switches it on by accident.

- Batch Upgrade (group I). Put your keeper drafts in a folder, set Batch count to the number of pictures, press Run once. Each picture is hired with its own saved prompt, so every character keeps its details. It skips re-making the draft, so it runs about 25 to 30 seconds a picture instead of 45 to 50.

- Send to keepers. Hover a draft (in the feed or on a Save node) and press K, or right-click a Save node and choose Send to keepers. It copies the picture into Anima/keepers for the batch. Change the key under Settings, Keybinding, search "Anima".

- Jump keys. Key 2 is models, prompts and sizes, 3 is Batch Upgrade; 4 is Regional Prompting, 5 is Image To Image, 6 is Image to Video; and 7 is Adetailer with its Load Image. The rest are Upscale, Inpainting and the others. You can also click the bookmark tab at the top to see every jump point.

- Regional prompt masks are in the download again (five of them), with the install note.

Install

1. Copy the whole anima_batch_loader folder into ComfyUI/custom_nodes/, restart ComfyUI, and refresh the browser tab. It is only needed for Batch Upgrade and Send to keepers. The rest of the workflow runs without it.

2. Open the v2.0 workflow.

3. For regional prompting, copy the default mask (Anima_Mask_3_Regions_1792x768.png) into ComfyUI/input/.

Good to know

- Batch Upgrade only reads PNGs made by this workflow, because each prompt is read from the picture.

- Set the index box to 0 before every batch and empty the keepers folder afterward.

- In Nodes 2.0 the yes/no state on the switch boxes does not redraw (a limitation of the rgthree switch nodes). They still work. Send to keepers works in both views, but the right-click item may only appear in the regular view. Hover and press K works in either.