Cartoon 3D CG is a custom checkpoint designed for cinematic, CG‑style cartoon renders with polished detail and dramatic atmosphere. It emphasizes material realism, expressive lighting, and strong composition, making it ideal for fantasy storytelling, epic battles, magical portraits, and stylized character design. Results often feel like frames from a CGI cutscene or animated feature film.

It excels at character‑driven fantasy, lush natural environments, atmospheric portraits, and decorative or romantic scenes — all with vibrant colors, smooth transitions, and a cinematic tone.

🤌🏻 Best for:

• Epic fantasy battles and large‑scale environments

• Cinematic portraits and half‑body shots

• Magical or surreal storytelling

• Game‑inspired cutscenes and CG‑like renders

• Decorative, romantic, or botanical scenes

✨ Core Features:

• CG‑style rendering: polished surfaces, strong material definition (fabric, metal, water, glow)

• Atmosphere: excels at fog, mist, magical glow, moonlight, and dramatic shadows

• Scalability: equally effective for single characters and complex multi‑character scenes

• Color and mood: vibrant palettes, cinematic contrast, and emotional tone

☝ Tips:

• Start promt with the quality stack: masterpiece, best quality, amazing quality, 4k, very aesthetic, high resolution, ultra-detailed, absurd, newest, cartoon, toon, 3d animation

• Use a negative filter for purity: modern, recent, old, oldest, anime, graphic, text, painting, crayon, graphite, abstract, glitch, deformed, mutated, ugly, disfigured, long body, lowres, bad anatomy, bad hands, missing fingers, extra fingers, extra digits, fewer digits, cropped, very displeasing, (worst quality, bad quality:1.2), sketch, jpeg artifacts, signature, watermark, username, (censored, bar_censor, mosaic_censor:1.2), conjoined, bad ai-generated

• A1111 → produces more stylized, illustrative results. It is well suited for posters, artbooks, and fabulous frames.

• Forge → enhances depth, materiality, and atmosphere. The scenes look solid and cinematic, like stills from a CGI movie.

• Hi‑Res Fix and ADetailer are recommended for portraits and complex scenes.

• VAEs: sdxl_vae and xlVAEC_c9. xlVAEC_c9 is just brighter

• Samplers and Schedulers: Euler a Automatic

• Steps: 20-35

• CFG: 5

• Width and Height: 768×1344, 832×1216

🩵 All examples are made in SD Forge
You can use this site to better understand the settings of the examples 🩵

I hope you enjoy! Follow for more, leave comments, put reactions and create!

You can help me with it, tips are always welcome 💌

1. Purpose & Nature of the Model
This model is provided strictly for research, artistic exploration, and experimental purposes, including NSFW content generation. It does not intend to copy, reproduce, or replace any existing artworks, characters, brands, or visual styles — including those protected by copyright, trademark, or patent law. All training and merging processes were performed solely for transformative learning of visual features, without direct reproduction of any specific protected works. All depicted characters are fictional and imagined as adults (21+); any resemblance to real persons, names, or events is purely coincidental and unintentional.
2. Liability & Responsibility
The author is not liable for any interpretations, generations, modifications, forks, merges, or downstream uses of this model by third parties. Users are solely responsible for ensuring their use complies with local laws, ethical standards, and platform policies, including those governing NSFW or adult content.
3. Warranty Disclaimer
This model is provided “as is”, without any express or implied warranties regarding accuracy, safety, completeness, or fitness for any particular purpose. While the author strives for quality, consistency, and responsible sharing, they assume no responsibility for any direct, indirect, incidental, or consequential damages arising from its use.
4. Redistribution & Derivatives
This model may be merged, adapted, or used as a base for personal, creative, or research purposes, provided that all such use remains non-commercial and is not presented as the original model. However, re-uploading, rehosting, or redistributing this specific model file — on external platforms (including but not limited to SeaArt, PixAI, Tensor.art, HuggingFace, or Civitai re-uploads) — is strictly prohibited without the author’s explicit permission. Derivative or merged versions created for personal or research use are permitted, as long as they are not presented as the original model. This restriction includes any rehosting, mirroring, or derivative listings intended to present the model as a separate publication. Any profiles, publications, or channels not affiliated with the official PupaBoo account on Civitai are considered unauthorized. The author shares this model exclusively for research and creative purposes on the official Civitai page.
5. License
This model is a derivative merge built upon architectures and components licensed under the following frameworks: SDXL, Illustrious, NoobAI-XL.

As such, this model remains subject to the terms and conditions of the above licenses, including the Open RAIL-M restrictions on harmful or commercial misuse, and the FAIPL-SD provisions for public AI sharing and research.