AllImageAI
By BaiduActive

ERNIE Image AI Generator

Baidu’s multilingual generator for Chinese, English and Japanese prompts with classic diffusion controls.

Ask AI for better images

Generate better images

Prompt ideas, settings and fixes for this model.

If the prompt isn't filled in, copy and paste it into your chat.

Prompt examples Verified
Original AI-generated editorial artwork for ERNIE Image
AI-generated showcase artwork for ERNIE Image. Editorial cover, not a benchmark result.
Editorial cover

Model Version

Mode

Output

ERNIE Image

Baidu’s multilingual generator for Chinese, English and Japanese prompts with classic diffusion controls.

BalancedUp to 2K1:116:99:164:3

Try one of these prompts

Model showcase

Curated outputs with their exact prompts and model provenance.

Developer
Baidu
Model
ERNIE Image
Status
Active
Max Output
2K
Input
Text
Recorded event
April 15, 2026 · Model announced
Last Verified
August 25, 2026
View release history and date evidence ↗

What Is ERNIE Image?

ERNIE Image is Baidu’s multilingual text-to-image model. Its hosted operation accepts Chinese, English and Japanese prompts and exposes negative prompts, guidance, step count and custom output dimensions.

ERNIE Image Model Versions

ERNIE Image Turbo

8 steps · 2K

From 1 Credits

Open this version →

ERNIE Image

Multilingual · 2K

From 3 Credits

ERNIE Image Turbo

Baidu’s faster ERNIE image tier for English, Chinese and Japanese prompts with optional expansion.

  • Resolutions: 1K / 2K
  • Reference images: not supported
  • Modes: Text to Image, Typography
  • Speed: Fast

ERNIE Image

Baidu’s multilingual generator for Chinese, English and Japanese prompts with classic diffusion controls.

  • Resolutions: 1K / 2K
  • Reference images: not supported
  • Modes: Text to Image, Typography
  • Speed: Balanced

What Is ERNIE Image Best For?

  • Chinese-language prompts
  • Japanese and English concepts
  • CJK poster drafts
  • Classic diffusion control

How to Use ERNIE Image

  1. Write in the language that best expresses the brief.
  2. Quote exact display text.
  3. Iterate at 1K.
  4. Use 2K only for the selected composition.

ERNIE Image Settings Explained

Negative Prompt

Exclude unwanted extra copy, blur, watermarks or duplicate objects.

Guidance

The documented default is 5; reduce it for more natural variation.

The panel above shows only the parameters ERNIE Image actually supports. Switch versions and unsupported controls disappear rather than sitting there doing nothing.

ERNIE Image Prompt Examples

Chinese poster

现代旅行海报,准确标题“山海之间”,青绿色海岸与雾中群山,简洁网格,留白充足,高级纸张质感

Japanese shopfront

雨の夜の小さな喫茶店、正確な看板「月灯り珈琲」、暖かな窓の光、映画的な写真

ERNIE Image Test Results

Testing in progress

Every model runs the same eight-category prompt set — portrait, product, poster text, anime character, interior, multi-reference, image editing and character consistency — recording generation time, success rate, text accuracy and identity retention. That run has not finished for ERNIE Image, so this page shows no scores. We do not publish numbers we have not measured.

ERNIE Image Limitations

  • Text-to-image only on this route
  • 2K is billed at four times the 1K pixel cost
  • Automatic prompt expansion is disabled for literal control

ERNIE Image Pricing and Credits

Credits scale with resolution and image count. The Generate button always shows what this run costs.

1K

3 credits / image

2K

9 credits / image

Failed generations cost 0 credits.

Commercial Use and Image Rights

Commercial use is subject to Baidu's model and service terms.

  • Model used: ERNIE Image
  • Model developer: Baidu
  • The inference route is managed internally and may change without changing the selected model
  • Uploads are sent only to the configured inference service needed to complete the generation
  • You can delete any generation from your history at any time

Frequently Asked Questions

Which prompt languages are supported?+

The model page explicitly lists English, Chinese and Japanese.

Why is prompt expansion off?+

It keeps exact wording under your control and avoids silently rewriting multilingual prompts.

Release History and Sources

April 15, 2026 · Model announced

This date applies to the event shown. Later API access, weights and availability on All Image AI may have different dates.

Explore this event on the image model timeline →