π Iβm preparing documentation on creating models using unique methods.
π Active users of my models will receive it for free.
Follow and share your creations.
π¨ v3 is a powerful tool for creating comics, graphic novels, and illustration work.
This version is designed for experienced users and professionals: it supports a wide range of techniques, styles, and detail levels. Inpainting is included.
β Recommended Negative Prompt
low detailed hands, (glowing eyes:1.2), (title, text:1.3), (title magazine, magazine cover, cover, picture-in-picture, multi frame, sequential, additional mini-scenes:1.3), speech bubble, thought bubble, text on image, dialogue box, word balloon, sound effects, frame, border, nudity, logo, hieroglyph, trade marksOptional additions if you want to further restrict the style:
realistic, 3d, cgi, volumetricShort or confusing prompts can create a page with a frame breakdown; this problem can be solved by increasing the emphasis on the key object in your prompt. Or change the "Sampling method" to DDIM.
βοΈ Recommended Settings
Sampling method: DPM++ 2M
Schedule type: Beta
Sampling steps: ~40
CFG Scale: 5β7
Size: 512x640px - safe size, max 832px and apscale.
π§ͺ Interesting Token Combinations (for flat 2D styles)
(sharp volumetric edges:1.3)(vibrant volumetric colors:1.3)(realistic:-1.1)Using this model, you'll discover many more amazing combinations.
π βTailβ Styles for Your Prompt
All of these can be freely mixed with each other.
Additions to your main prompt:
, (modern comic style), sharp inking, vivid cel-shading, (ink volumetric:1.25), crisp outlines, dynamic highlights, subtle rim light, (Manga style:1.25), color lineart, screentone shading, sharp focus on eyes, speed lines, high energy pose, (vibrant volumetric colors:1.3), subtle digital glow, sharp edges, bright, reflections, glare, bokeh, modern comic style, sharp inking, vivid cel-shading, crisp outlines, dynamic highlights, subtle rim light, volumetric accents, flat color shading, cel shading, clean lineart, vibrant comic book palette, solid color fills, sharp edges, dynamic pose, (comic book style:1.15), (bold lineart:1.25), (variable lines:1.2), (bold lineart:1.25), (variable lines:1.2), (dynamic lines:1.2), (tapered lines:1.2), (Ink Wash technique:1.25), delicate brushstrokes, expressive movement, focus on atmosphere, (vibrant volumetric colors:1.15), subtle digital glow, (sharp volumetric edges:1.15), bright, reflections, glare, bokeh, (volumetric realistic:1.1), (Manga style:1.25), (cell shade:1.3), color lineart, screentone shading, sharp focus on eyes, speed lines, high energy pose, (vibrant flatten colors:1.1), (sharp flat edges:1.1), bright, reflections, glare, bokeh, (Manga style:1.25), color lineart, screentone shading, sharp focus on eyes, speed lines, high energy pose, (vibrant volumetric colors:1.3), subtle digital glow, sharp edges, bright, reflections, glare, bokeh, (Manga style:1.1), color lineart, screentone shading, sharp focus on eyes, speed lines, high energy pose, bright, glare, reflections, (Modern comic style:1.2), detailed rendering, vibrant colors, subtle digital glow, sharp edges, (Ripped edge style:1.1), distressed paper texture, halftone printing, gritty texture, noir aesthetic, stylized shadows, (classic American comic style:1.1), thick outlines, clean color fill, flat shading, vibrant primary colors, (Pop Art style:1.2), bold color blocks, Ben-Day dots texture, heavy outline, portrait of a hero, Roy Lichtenstein inspired, old paper, (retro comic style:1.15), (comic book style:1.15), (bold lineart:1.25), (variable lines:1.2), (old comic book colors, retro comic style:1.2), (Chinese comic style:1.25), (Ukiyo-e Inspired style:1.25), (Shojo Manga style:1.25) , (Guofeng style:1.25), (Guofeng aesthetic:1.3), (Wuxia Aesthetic:1.3)π Reference Material
Prepared this table for personal use, but it may help you as well:
Link: Master Test
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Link: Reaction of 22 models to the "Children's illustration style" token (3456x11264 px).
"BRAVO - True Comics v1" - Comic Art Style Model β Flexible, Sharp, Stylized
β οΈExcellent results with:
Schedule type: Beta
Sampling method: Euler a, DPM++ 2M, DDIM, Heun, DPM++ SDE
CFG Scale: 5-7, Sampling steps: 40+
This model is designed for generating illustrations and scenes in various comic book styles. The dataset was carefully curated to remove panel frames and text bubbles, focusing only on clean character and scene illustrations.
Itβs also well-suited for creating storyboards and visual concepts for video production and film pre-production, offering consistent framing, strong silhouettes, and stylized lighting that support narrative planning.
Beyond traditional comic formats, the model handles hybrid and unconventional styles with confidence β making it a powerful tool not only for comic creation, but for art direction in game development, motion graphics, or even experimental illustration projects.
The model retains style without needing βcomicβ in the prompt and responds well to prompt adjustments. It's excellent for training LoRA on comic datasets due to its unburnt weights and balanced detail retention.
π‘ For better color harmony, try adding "color scheme" to your positive prompt.
π¨ Supported Styles (you can guide the output via prompts):
Classic American Comics (Marvel/DC)
Dynamic poses, bold inking, halftone shading, superhero action scenes.European Comics (Tintin, Moebius)
Ligne claire style, flat clean colors, thin lines, detailed environments.Manga (Black & White, Action-Focused)
Speed lines, screentones, ink wash, high contrast B/W, expressive strokes.Graphic Novel (Dark Realism, Sin City)
Moody lighting, noir textures, chiaroscuro, dramatic realism.Vintage Comic (Golden Age, 40s-50s)
Dot textures, bold colors, Jack Kirby-style action, old print aesthetics.Cartoony / Satirical Style (Mad Magazine)
Exaggerated expressions, wobbly ink lines, messy watercolor look.Modern Webtoon / Digital Comic
Clean cel shading, thick outlines, soft gradients, cozy slice-of-life scenes.
π οΈ Use this model for:
Comic-style concept art
Stylized inpainting corrections (via matching inpaint model)
Creative storytelling assets
Stylized game assets or character sheets
πΌοΈ Recommended settings:
Resolution: 512Γ512 (safe), up to 640Γ768 (with caution)
SD VAE vae-ft-mse-840000-ema-pruned
Hi-Res Fix: ESRGAN_4x
Negative Prompt:
frame, bad hands, extra finger, bad anatomy
π§ Inpainting support:
If generation goes off-style or needs cleanup, use the inpainting variant of this model to make precise edits within the same aesthetic.
Enjoyed the model? Drop a couple of buzz β you'll look cool, and I'll be grateful β‘π

