The simplest way to self-host Dreamshaper. Launch a dedicated cloud GPU server running Lab Station OS to download and serve the model using any compatible app or framework.
Direct Download
Download model weights for local inference. Must be used with a compatible app, notebook, or codebase. May run slowly, or not work at all, depending on your system resources, particularly GPU(s) and available VRAM.
Forge is a platform built on top of Stable Diffusion WebUI to make development easier, optimize resource management, speed up inference, and study experimental features.
Train your own LoRAs and finetunes for Stable Diffusion and Flux using this popular GUI for the Kohya trainers.
Model Report
Lykon / Dreamshaper
DreamShaper is a Stable Diffusion model optimized for diverse artistic styles, with particular strength in illustrations and character generation. Version 8 builds on previous iterations' improvements in lifelike elements and anime-style imagery, while offering LCM variants for faster processing.
Explore the Future of AI
Your server, your data, under your control
DreamShaper is a versatile Stable Diffusion checkpoint model created by Lykon, designed as an open-source alternative to Midjourney. The latest version (v8) was released on October 4, 2024, building upon previous iterations with significant improvements in character LoRA handling, photorealism, and NSFW content generation capabilities.
Model Variants and Evolution
DreamShaper is available in several variants, each optimized for different use cases. The LCM (Latent Couple Model) versions prioritize speed for real-time applications, though they may sacrifice some image quality and require specific LCM samplers. Other variants feature different VAEs (Variational Autoencoders) and pruning levels to suit various needs.
The model's evolution shows interesting specialization across versions:
Version 4: Excelled in anime generation with less reliance on LoRA networks
Version 5: Notable for photorealism capabilities
Version 7: Improved LoRA support, NSFW content, and general realism
Version 8: Enhanced character LoRA handling and photorealism
Technical Capabilities and Usage
DreamShaper demonstrates particular strength in generating:
Artistic illustrations
Anime-style images
Landscape compositions
Photorealistic renders
For optimal results, users should:
Use highres.fix or img2img for higher resolutions
Avoid excessive use of "restore faces" or ADetailer, as these can homogenize facial features
Consider specific variants based on use case (e.g., LCM for speed, inpainting variants for touch-ups)
Licensing and Legal Considerations
DreamShaper is licensed under CreativeML Open RAIL-M with an additional addendum. The model maintains open-source principles while establishing clear usage guidelines through its licensing structure.
