Naxos
Fast generations
CFG 1 - 2
Steps 10 - 30
V7
Peak Comics This mix is my latest gem!
BK-Nora
Euler_A - Heun - Deis (if generated on site)
CFG 5
Steps 35 and above
Test at will and share your creations. As always I will be gifting buzz to the ones I enjoy :D
`````````````````````````````````````````````````
V6 Oberon Version
Uses my Oberon modelmerge to add more linework and comic influence
I tested with a new sampler which runs at low CFG without a DMD Lora
I am experimenting to provide both normal steps/CFG and viable fast generations without overcooking.
Boost V1
~
This merge is pre-baked and curated with a specific aesthetic in mind. To ensure high-fidelity outputs and avoid "frying" the image details, please adhere to the following settings.
🔹 Generation Parameters
CFG / Guidance Scale: Keep it low, between 4 - 6.
Steps: 20 - 45 steps are sufficient for convergence.
🔹 LoRA & Style Guidance Please avoid stacking multiple LoRAs. Because this model already has a strong, baked-in style, stacking heavy style LoRAs will likely degrade the output quality.
Stabilizers: Not necessary. The model is stable on its own.
If you must use a LoRA: Keep the weight low (0.4 to 0.8) and use only one at a time for best results.
↓↓V2↓↓
MaturaComix — V2 (oneSHOT) | Next-Gen Hybrid Toon-Comix Style 🔥
V2 (oneSHOT) pushes the signature MaturaComix look to a whole new level — sharper, cleaner, more vibrant, and more dynamic than ever. This version focuses on delivering high-impact illustrations that feel like a perfect fusion between premium anime aesthetics and Western comic book boldness.
✨ What’s New in V2?
Upgraded Merge Stack: Rebuilt using newer, higher-quality models for cleaner linework, smoother gradients, and richer color harmony.
Improved Consistency: More stable anatomy, better facial fidelity, and dramatically reduced color bleeding.
Balanced Shading & Tone: Enhanced contrast and polished highlights for a glossy, finished-art look.
Optimized for Modern Samplers: Especially tuned for dpmpp_3m_sde_dynamic_eta + kl_optimal, giving razor-sharp detail without artifacts.
🎨 Core Style Features
Hybrid Anime × Comic Fusion: Soft and expressive anime faces combined with bold, structured Western-comix anatomy.
Mature & Stylized Appeal: Powerful silhouettes, dynamic posing, and an illustrative look that pops instantly.
Vibrant, Polished Rendering: Smooth color transitions, glossy accents, and a clean, professional-grade finish.
Expressive Lineweight: A crisp balance of thick contours and delicate interior details for a premium illustrated feel.
🧬 Model Lineage
Built from a refined blend of:
WAI-Mature-Illustrious
Nova Mature XL
NeroMatura
Plus specialty LoRAs to add that unmistakable “comic seasoning” in every render.
V2 enhances and stabilizes all of these inputs for next-level output quality.
⚙️ Recommended Settings
Sampler:
dpmpp_3m_sde_dynamic_eta + kl_optimalCFG: 3 – 6
Steps: 20 – 50
Designed for fast, clean, one-shot renders without needing heavy prompting.
💥 Perfect For
Character-focused illustrations
Anime/comic hybrid art
Glossy stylized pinups
High-impact posters and cover-style imagery
Artists looking for fast, high-quality, ready-to-use results
Description
Finer Details + More Anime-esque
Leaner outlines
Recommended sampler x Scheduler
dpmpp_3m_sde_dynamic_eta + Kl_Optimal
5 CFG max, 20 - 60 steps range
FAQ
Comments (4)
I usually generate using SwarmUI locally. I notice your sampler is quite peculiar. Any other suggestion, or does the model work only with that one?
I use comfyui this Sampler is of the most recent ones. For this type of images "anime/comics" I with solid linework and stuff it is considered to be one of the best in terms of quality. But any other of the normal ones are fine euler a, but dpmpp 3m sde etc.
I am currently testing kl_optimal scheduler and on this model it's better than karras, beta and sgm_uniform in my opinion











