Z-Image-Omni-Base
Strategic evolution of Z-Image-Base, featuring omni pre-training for unified image generation and editing, avoiding complexity and performance loss of task switching.
Overview
Z-Image-Omni-Base marks a strategic shift from the original 'Base' model towards an 'omni' (omnipotent) pre-training architecture. It unifies image generation and editing/inpainting tasks within a single framework using the Scalable Single-Stream Diffusion Transformer (S3-DiT). This omni pre-training allows for seamless transitions between generating new images and editing existing ones without the need for separate specialized models, offering higher parameter efficiency and flexibility for developers.
Features
- Omni Pre-training for Unified Generation and Editing
- Seamless Task Switching without Performance Loss
- Scalable Single-Stream DiT (S3-DiT) Architecture
- Supports Cross-Task LoRA Adapters
- 6B Parameter Efficiency
- Superior Performance on Complex Multimodal Tasks
Images
Related Links
Popular Tools
Explore our most popular creative tools
Z-Image Edit
Upload image, transform with one sentence
Creative Engine
One sentence, AI provides infinite prompt creativity.
Image Parse
Upload image, retrieve prompt instantly.
Prompt Library
Discover thousands of high-quality AI prompts.
Style Library
Explore curated artistic styles for your creations.
Z-Video Generator
Generate creative videos from text or images with AI.
AI Image Generator
Turn your text into stunning images instantly.