A VAE (Variational Autoencoder) model for SDXL developed by Stability AI. Improves latent space encoding and decoding for image generation.
Parameters
Undisclosed
Context Window
License
Proprietary
Release Date
2023-06-21
API Pricing
API pricing for this model is not yet available
Strengths
- •Optimized for SDXL architecture
- •Produces high-quality latent representations
- •Efficient encoding and decoding
Weaknesses
- •Only compatible with SDXL models
- •Not a standalone image generator
- •Limited to latent space operations
Use Cases
- •Image encoding to latent space
- •Latent-to-image decoding for SDXL
- •Improving SDXL generation quality