ERNIE-Image is Baidu’s open-weight text-to-image model designed for accurate in-image text, complex prompt following, multi-panel layouts, and bilingual Chinese-English generation for posters, comics, infographics, and text-rich visuals.
Sign in to generate multiple images at once






Explore ERNIE-Image examples for posters, comics, infographics, bilingual visuals, multi-panel storyboards, and complex scene layouts.

Great for campaign posters, brand visuals, and social covers with clearer headline hierarchy and in-image text.

Designed for comic strips, speech bubbles, and multi-panel storytelling with stronger layout consistency.

Useful for labeled visuals, explainers, and structured infographic layouts with readable section text.

Built for posters, signage, and covers created from Chinese, English, or mixed bilingual prompts.

Great for storyboards, sequential scenes, and structured compositions that need stronger panel control.

Suitable for multi-object, multi-character scenes where composition and prompt detail both matter.
ERNIE-Image is especially useful for text-rich, structured, and bilingual visual creation.
Create promo posters, campaign graphics, event visuals, and social covers with stronger text readability and layout control.
Generate comic dialogue scenes, storyboards, and narrative layouts with stronger multi-panel structure.
Useful for labeled diagrams, infographic-style layouts, and explanatory visual content.
Ideal for creative work using Chinese, English, or mixed prompts for bilingual visual communication.
Create structured visuals in four simple steps with the existing online generator.
Select ERNIE-Image or ERNIE-Image Turbo based on the kind of visual result you want.
Describe the poster, comic, infographic, or text-rich image you want, including layout, wording, and style.
Generate your image and review the text rendering, composition, and overall visual structure.
Adjust your prompt, refine the result, and export the version that best fits your publishing or design needs.
ERNIE-Image focuses on text-rich, structured, and bilingual image generation tasks.
Better for poster copy, labels, comic dialogue, and text-heavy visuals where many image models still struggle.
Handles scenes with multiple objects, spatial relations, constraints, and detailed instructions more consistently.
Especially useful for visual tasks that need stronger layout control, structure, and text-image coordination.
Works well with Chinese, English, and mixed prompts for multilingual creative work.
Core ERNIE-Image capabilities for posters, comics, infographics, and text-rich visuals.
Better for images that need readable titles, labels, dialogue, and dense text content.
Useful for prompts with multiple objects, attributes, relationships, and structured composition.
Designed for posters, comics, storyboards, and other structured visual formats.
Supports Chinese, English, and mixed bilingual prompt input.
Start generating instantly and try ERNIE-Image without creating an account.
Built around Baidu’s open-weight text-to-image model for users who want to explore ERNIE-Image online.
Answers to common questions about using ERNIE-Image online.
No login required. Start generating posters, comics, infographics, and text-rich visuals right away.
This website is an independent AI image generation service and is not affiliated with Baidu. References to ERNIE-Image are used to describe the supported or showcased model capability.