Downloads · 30 days
160
22% of all-time downloads
muverqqw/DreamCoil-V1.0
DreamCoil-V1.0 is a text-to-image model from muverqqw. Use it when you need an image from a text prompt. It is set up for diffusers. The card lists the license as creativeml-openrail-m.
<div style="background: radial-gradient(circle at top, 1a0b2e 0%, 000000 100%); border-radius: 20px; padding: 40px 20px; text-align: center; border: 1px solid rgba(255,255,255,0.1); box-shadow: 0 0 40px rgba(76, 29, 1…
Downloads · 30 days
160
22% of all-time downloads
All-time downloads
717
Public
Repo size
76.4 GB
Likes
0
Public
Click a slice to open those files.
.safetensors55.5 GB · 100%
From the Hugging Face model README
After a period of rigorous optimization and development, DreamCoil is back!
We are thrilled to announce the release of DreamCoil-V1.0. All classic DreamCoil versions have received substantial internal updates, delivering significantly improved coherence, faster generation, and better artistic consistency across various styles.
Crucially, this release introduces our new, high-quality flagship lineup: Lumin. This series is built for superior detail, complex lighting, and style adaptability, and is set to replace the older 'PyroNet' and 'masterpiece' lines starting with DreamCoil-V1.1.
| Field | Value |
|---|---|
| Release Year | 2025 |
| Main Developer | @IceL1ghtning |
| Base Models | stabilityai/stable-diffusion-xl-base-1.0 and OnomaAIResearch/Illustrious-XL-v1.0 |
| Model Size | 6.9 GB (Estimated) |
| Training Language | English (en) |
| Library | diffusers |
| License | CreativeML Open RAIL-M |
The Lumina series focuses on high-fidelity, polished output. Each model is specialized for a distinct aesthetic, ensuring top-tier results in its domain.
Lumina-Noob serves as a robust foundational model, derived from a leaked checkpoint of the highly capable Noob-AI model. It provides exceptional versatility, excelling in photorealistic detail, clean lines, and high-contrast visuals, making it perfect for detailed character sheets and renders.
| Aspect | Details |
|---|---|
| Style Focus | Detailed Semi-Realism, Clean Anime Rendering |
| Source | Leaked Noob-AI Checkpoint |
| Preview Image | ![]() |
The core model of the Lumina series, serving as the DreamCoil flagship for general-purpose high-quality anime and artistic illustration. It excels at complex, narrative-driven compositions, dynamic lighting, and rendering intricate visual effects like fire, water, and fireworks.
| Aspect | Details |
|---|---|
| Style Focus | High-Quality Anime Illustration, Complex Compositions |
| Aesthetic | Vibrant colors, high detail, atmospheric effects |
| Preview Image | ![]() |
This model is meticulously tuned for generating expressive characters in a distinct, clean Cartoon style, specifically optimized for Western and contemporary Asian cartoon aesthetics. It favors bold, clean lines, clear cell-shading, and strong visual storytelling, making it ideal for profile pictures and expressive scenes.
| Aspect | Details |
|---|---|
| Style Focus | Modern Cartoon, Expressive Character Art |
| Aesthetic | Bold lines, clear shading, vibrant color blocks |
| Preview Image | ![]() |
Designed specifically for generating top-tier, high-resolution Pixel Art. The Lumina (Pixel) model provides consistent 2D sprites, dynamic scenes, and excellent character anatomy while faithfully adhering to pixel grid limitations. It perfectly captures the detailed, colorful aesthetic of modern retro games.
| Aspect | Details |
|---|---|
| Style Focus | High-Resolution Pixel Art, Detailed Retro Gaming Aesthetic |
| Consistency | Excellent adherence to pixelation and color palettes |
| Preview Image |
The future of DreamCoil is focused entirely on refining and expanding the Lumina architecture.
PyroNet and masterpiece models.We extend our sincere thanks to everyone who contributed to this monumental update.
@IceL1ghtningNoob-AI (for the foundation and inspiration of Lumina-Noob)Katya Mikhaylets, Misha Kashin, Vadim Galyareta & Polishuk Artem