モデルシリーズERNIE Image
ERNIE-Image is an open-weight text-to-image model developed by the ERNIE-Image Team at Baidu, built on a single-stream Diffusion Transformer (DiT) with 8B parameters and paired with a lightweight Prompt Enhancer that rewrites short prompts into richer, more structured descriptions before passing them to the diffusion backbone. NYU Shanghai RITS Released on April 15, 2026 under the Apache 2.0 license, it transforms natural language descriptions into detailed imagery with particular strength in text rendering and structured layout generation. ERNIE-Image is designed not only for strong visual quality, but for controllability in practical generation scenarios where accurate content realization matters as much as aesthetics — making it well-suited for commercial posters, comics, multi-panel layouts, and other content creation tasks that require both visual quality and precise control.
探索