Autoplay
Run Locally
SumiInk_h3_v2.safetensors
148 MB · SafeTensor · bf16
Downloads are served by Civitai. Some creators ask for a Civitai login first.
About This Model
MINIMAX H3 · LoRAs
by RisingV - SumiInk - Sumi-e and Ink wash painting inspired style [MiniMax H3] - A style LoRA for creating a rough ink brush painting style inspired by Sumi-e and other east asian ink wash painting styles.
Creator notes
A style LoRA for creating a rough ink brush painting style inspired by Sumi-e and other east asian ink wash painting styles. Thank you to the people helping me test the H3 version, especially to who also made the showcase videos for me. Check out their ! I couldn't test it myself but it seems to work best starting the prompt with "A sumiink painting style animation". For more complex prompts you maybe need additional style description or use prompt enhancement for the style. Using a Turbo LoRA seems to decrease the visual quality and prompt adherence. In my tests with Minimax H3 I found you should use at least 30 steps without Turbo LoRA to get a high quality video. Therefore I remade two of the showcase videos without using a Turbo LoRA. Thank you to and for using their LoRAs and to for using some prompt remixes in the showcase images of the Krea 2 version.
trigger words H3: A sumiink painting style animation
trigger words Krea 2: SumiInk, traditional media
If you like the model please leave some reactions on the showcase images or videos. Also I would like to see the images or videos you generated. So if you really like the model please click "Add Review", leave a like and add some images, so I get notified you did. If you like you can also write a short review.
Trigger words
A sumiink painting style animation
SHA256
80A18A354C66DCBEF5B80F1641386D969182BDC88180858EC5B6BACBFD3844DD
ModelTrends.ai Model ID
#27282
Flag content
Example renders
Prompts and settings shared by the people who made these renders on Civitai. Pick a render to see how it was made.
Settings
Steps
30
Seed
1002643565
Size
704x960
Weight used
1
Prompt
integrated_multimodal_description: [Shot 1] Sumi-ink painting style animation, 2D-animated ink-wash short. Rice-paper ground, warm off-white, visible paper tooth. Forms built from wet grey washes and dry black brush calligraphy. A young adult woman, fully clothed: long loose white shirt, long blue jeans, black long hair in twintails, black eyes. Medium shot, three-quarter view, she stands in an empty wash-space with her left hand on a sheathed katana at her hip. The camera arcs around her with large amplitude at fast speed. She grips the hilt, draws the blade in one clean upward stroke, ink splash along the steel edge. She inhales. A sharp young adult female voice (S1) shouts: <d>[Japanese] せいっ!</d> [Shot 2] At 00:05.000, the camera cuts to a low wide angle and pushes in with large amplitude at fast speed as she leaves the ground, both knees tucked, katana raised over her right shoulder, twintails flung back in two dry-brush ribbons. She stays fully clothed. She travels toward camera through empty paper space and holds the strike pose as the wash settles.
overall_soundscape: Dry cloth snap from the shirt. Steel whisper as the blade leaves the sheath. A short grunt on the takeoff. Air rush under her jump. No crowd. No extra voices after the shout.
non_diegetic_music: N/A
Settings
Sampler
Euler
Steps
8
Size
896x1152
Prompt
For the target video, at 0.00 seconds into the target video, <Picture 1> (from [Shot 1]) is fully referenced.
integrated_multimodal_description: [Shot 1] Sumi-ink painting style animation. The woman in <Picture 1> starts in the same over-shoulder pose, rain still falling as thin black-grey ink strokes on empty paper. She slowly lifts her chin and looks upward into the rain. After a short hold she lets go of the katana with one hand; the other hand keeps the sword. She raises that free palm, open to the sky, and a few ink-rain drops land on the skin. Her mouth shifts into a small, reserved smile. She then brings that same hand up and takes the umbrella handle. She turns her back to the camera in one unhurried pivot. The umbrella tilts down and hides her head; only the dark kimono, the hair bow, and the long black tail of hair remain visible. She walks slowly away into the distance, shrinking in frame. As she recedes the camera pulls out with small amplitude at slow speed, staying behind her. The paper around her is empty at first. Then the falling black-grey watercolor ink rain strikes the ground and the air, and old Japanese wooden houses appear where the drops land — roofs, eaves, and lattice walls washing into view as if they had been invisible and the ink rain is painting them onto the paper. The houses stay; they do not flicker or redraw. She keeps walking between them until the last beat. No cut. No other people. No readable text.
overall_soundscape: Rain only. Raindrops on the umbrella, on her open palm, and on the ground. Soft cloth as she turns. No voices. No sword clinks. No crowd.
non_diegetic_music: N/A
Settings
Size
704x960
Prompt
A woman stands alone in purgatory. An eerie black hole stirs. Energy Like a black mist swirling. When the girl begins to walk quietly, It ends with a terrifying, demon-like face appearing from the hole. terrifying, terrifying.
japanese sumi artwork motion .
Settings
Steps
40
Seed
1747964041
Size
704x960
Weight used
1
Prompt
A sumiink painting style animation
We see a cherry tree in a beautiful asian landscape with distant mountains fog. A wind blows through, causing a few petals of flowers on the tree to come loose and gently blow away, the camera follows them as they drift down to a pond with lily pads and koi fish swimming in it, and one of the fish leaps out of the water with a splash as the petals land in the water and the video end.
Settings
Sampler
Euler
Steps
8
Size
896x1152
Prompt
integrated_multimodal_description: A SumiInk painting style animation. [Shot 1] A tight medium close-up, slight low angle. A young adult female red fox sits at a small wooden kitchen table, winter afternoon light from a window on the left. She wears a cream knit sweater, sleeves pushed to the elbows. Both paws rest around a white ceramic mug. Steam rises from the mug in a thin unhurried thread. She looks down into the cup for a beat, ears angled slightly forward. Then she lifts her chin toward the window, blinks once, and a low, warm adult female voice (S1) says: <d>[English] Still snowing.</d> Her mouth closes. She takes one small sip. The camera pushes in with small amplitude at slow speed toward her eyes and the mug rim. No other people. No readable text on the mug or in the room.
overall_soundscape: Quiet indoor room tone. Soft radiator tick. Distant muffled traffic through glass. A faint ceramic tap when the mug leaves the table and returns. Her breath after the sip.
non_diegetic_music: N/A
Settings
Sampler
Euler
Steps
8
Size
1184x864
Prompt
integrated_multimodal_description: [Shot 1] Sumi-ink painting style animation, 2D-animated ink-wash short. Rice-paper ground, warm off-white, visible paper tooth. The ocean is wet grey-blue washes with dry black calligraphy for foam edges. A small wooden ship sits mid-frame on the water, hull as flat inked planes and a few wet folds in the sail. Wide side view, horizon a single pale wash. No readable text. No people on deck. The camera holds a wide tracking shot with small amplitude at slow speed, sliding with the ship as large swells lift and drop it. The ship rides the water and is moved by the waves; it does not capsize. Some waves build high in the foreground and mid-distance, crest, and crash back into the sea without striking the hull. Spray is ink spatters that fall short of the ship. The vessel pitches and yaws with each swell and continues across the frame until the wash settles.
overall_soundscape: Open sea and wind only. Long rolling surf, water slapping the distant surface when a wave collapses away from the ship, steady wind across the paper space. No voices. No music source. No extra impacts on wood.
non_diegetic_music: N/A
Similar Models
AI Models for Image, Video & Text Generation
sumiink h3 is a LoRA for the MiniMax H3 family listed on ModelTrends.ai, a read-only catalog of open source AI image, video and text models. Compare it with other MiniMax H3 models, check its heat score to see how it is trending, and open it on PirateDiffusion or BitVector to try it.
Browse by Model Family
Text / LLM Models
411
Anima Models
509
Chroma Models
14
Flux Models
1,071
Flux 2 / Klein Models
1,237
MiniMax H3 Models
331
Hunyuan Models
197
Ideogram Models
20
Krea2 Models
674
Qwen Models
17
Qwen2 Models
36
Ltx2 Models
150
Stable Diffusion 1.5 Models
6,465
Stable Diffusion XL 1.0 Models
15,432
Zimage Models
804
Wan Models
919

Model
Trends
.ai
© ModelTrends.ai
|
Made in Japan
|
© 2026
