Tongyi Qianwen Text to Image Model Interface Document

1. Interface Basic Information

Model Name: qwen-image-plus (Qwen Image Generation Enhanced Version)

Base URL: https://llm.ai-nebula.com/v1/images/generations

Authentication Method: Bearer Token

Auth Token: Bearer sk-xxxxxxxxxx

Core Capabilities:

✅ Text-to-Image (Generate images from pure text descriptions)

✅ Chinese/English Text Rendering (Excels at generating complex text within images)

✅ Various Art Styles

✅ Intelligent Prompt Extension

❌ Does Not Support Image-to-Image

Supported Dimensions: 1328×1328, 1664×928, 928×1664, 1472×1140, 1140×1472

Billing Method: Per image billing (¥0.18/image, Official price ¥0.2/image)

Official Documentation: Alibaba Cloud Bailian Model Interface Documentation (qwen-image-plus)

1.1 Core Parameters Description

Parameter Name	Type	Required	Description	Example Value
model	string	Yes	Model name	qwen-image-plus
prompt	string	Yes	Text prompt	"A cute orange kitten"
parameters	object	Yes	Generation parameters	See description below

1.2 Request Parameter Structure

qwen-image-plus supports two parameter formats:

Format 1: Simplified Format (Recommended)

Directly flattened parameters, more concise:

{
 "model": "qwen-image-plus",
 "prompt": "Your prompt content",
 "parameters": {
  "size": "1328*1328",      // Image size (Required)
  "negative_prompt": "",     // Negative prompt (Optional)
  "prompt_extend": true,     // Prompt extension (Optional, default true)
  "watermark": true        // Watermark (Optional, default true)
 }
}

Format 2: Full Format

Using the full structure with extra and input.messages:

{
 "model": "qwen-image-plus",
 "prompt": "Your prompt content",
 "extra": {
  "input": {
   "messages": [
    {
     "role": "user",
     "content": [
      {
       "text": "Your prompt content"
      }
     ]
    }
   ]
  },
  "parameters": {
   "size": "1328*1328",
   "negative_prompt": "",
   "prompt_extend": true,
   "watermark": true
  }
 }
}

Suggestion: Use Format 1 (Simplified Format), as it is more concise and readable.

1.3 Supported Image Sizes

Size	Aspect Ratio	Usage Scenarios
1328×1328	1:1	Square images, avatars, social media
1664×928	16:9	Widescreen, video covers, desktop wallpapers
928×1664	9:16	Mobile vertical screen, short video covers
1472×1140	4:3	Landscape photos
1140×1472	3:4	Vertical photos, posters

2. Text-to-Image Functionality

2.1 Basic Text-to-Image (Square 1:1)

Generate a square image:

2.2 Widescreen Image Generation (16:9)

Generate a landscape image with a 16:9 widescreen ratio:

2.3 Vertical Image Generation (9:16)

Generate a poster suitable for a mobile vertical screen:

3. Advanced Features

3.1 Text Rendering (Core Advantage)

qwen-image-plus excels at generating complex text within images, supporting Chinese/English text, multi-line layout, and paragraph-level text generation:

Text Rendering Usage Scenarios:

Poster design (with titles and explanatory text)

Couplets, Calligraphy works

Product packaging design (with brand text)

Advertisement images (with slogans)

Educational charts (with labels)

3.2 Negative Prompt

Use negative prompts to exclude unwanted elements:

Common Negative Prompt Scenarios:

Exclude low quality: "blur, noise, low resolution, pixelated"

Exclude specific elements: "watermark, text, logo, copyright sign"

Exclude styles: "cartoon, anime, abstract, realistic"

Exclude defects: "distortion, deformation, extra limbs, wrong proportions"

3.3 Prompt Extension

Control whether to automatically extend and optimize prompts via the prompt_extend parameter:

Enable Prompt Extension (Recommended)

The system will automatically extend "City under sunset" into a more detailed description to improve the generation result.

Disable Prompt Extension

If your prompt is already very detailed, you can turn off extension:

Prompt Extension Suggestions:

✅ Short prompts: Enable extension (prompt_extend: true)

❌ Detailed prompts: Disable extension (prompt_extend: false)

✅ Need precise control: Disable extension

✅ Rapid creation: Enable extension

3.4 Watermark Control

Control whether to add a watermark to the generated image via the watermark parameter:

Note: Disabling watermarks may violate certain terms of use; please choose carefully.

4. Response Handling

4.1 Response Format

A successful response (Status Code 200) returns JSON containing image data:

{
 "code": 200,
 "msg": "Operation successful",
 "data": {
  "data": [
   {
    "url": "https://dashscope-result-wlcb-acdr-1.oss-cn-wulanchabu-acdr-1.aliyuncs.com/...",
    "b64_json": "",
    "revised_prompt": ""
   }
  ],
  "created": 1762411334,
  "metadata": {
   "output": {
    "choices": [
     {
      "finish_reason": "stop",
      "message": {
       "content": [
        {
         "image": "https://..."
        }
       ],
       "role": "assistant"
      }
     }
    ],
    "task_metric": {
     "FAILED": 0,
     "SUCCEEDED": 1,
     "TOTAL": 1
    }
   },
   "request_id": "c9c7658e-a78e-4f3d-b59c-c25ab83c833b",
   "usage": {
    "height": 1328,
    "image_count": 1,
    "width": 1328
   }
  }
 }
}

4.2 Response Field Description

Field Path	Type	Description
code	int	Response status code, 200 indicates success
msg	string	Response message
data.data[]	array	Array of generated images
data.data[].url	string	Image URL address (valid for approx. 24 hours)
data.created	int	Generation timestamp
data.metadata	object	Manufacturer's original response data
data.metadata.usage.image_count	int	Number of images generated
data.metadata.usage.width	int	Image width
data.metadata.usage.height	int	Image height

4.3 Error Handling

If the request fails, an error message will be returned:

{
 "code": 400,
 "msg": "Parameter Error: Unsupported image size",
 "data": null
}

Common Error Codes:

400: Parameter error (e.g., wrong model name, invalid size format, etc.)

401: Authentication failed (Invalid or expired API key)

403: Insufficient permissions or balance

429: Request too frequent, rate limit exceeded

500: Internal Server Error

5. Best Practices

5.1 Size Selection Suggestions

Social Media:

WeChat Moments: 1328×1328 (1:1)

Weibo Banner: 1664×928 (16:9)

Douyin (TikTok) Cover: 928×1664 (9:16)

Xiaohongshu: 1140×1472 (3:4)

Design Usage:

Website Banner: 1664×928 (16:9)

Poster: 1140×1472 (3:4) or 928×1664 (9:16)

Product Image: 1328×1328 (1:1) or 1472×1140 (4:3)

Phone Wallpaper: 928×1664 (9:16)

Video Related:

Video Cover: 1664×928 (16:9)

Short Video Cover: 928×1664 (9:16)

5.2 Prompt Optimization Suggestions

Text Rendering Prompt Tips

The core advantage of qwen-image-plus is complex text rendering. Clearly state the text content in the prompt:

✅ Good Prompt (Clear text content):

"A poster with the title 'Summer Sale' and subtitle '50% off site-wide', background is a fresh beach scene"

❌ Bad Prompt (Vague description):

"A promotional poster"

✅ Good Prompt (Detailed text layout):

"A Chinese couplet, left scroll 'Spring returns to the earth, thousands of mountains are beautiful', right scroll 'Sun shines on the divine land, all industries prosper', horizontal scroll 'Everything is renewed', background is red Xuan paper, font is regular script in gold"

❌ Bad Prompt (Missing text):

"A Chinese couplet"

General Prompt Tips

Clarify Composition Direction:

Horizontal: Use "horizontal composition", "widescreen view"

Vertical: Use "vertical composition", "vertical view"

Add High-Quality Keywords:

"High quality", "HD", "Professional photography"

"8k resolution", "Rich details"

Specify Art Style:

"Realistic style", "Oil painting style", "Watercolor"

"Chinese style", "Cyberpunk", "Modern minimalist"

Describe Lighting and Atmosphere:

"Soft lighting", "Dramatic light and shadow"

"Warm tones", "Cool tones"

5.3 Negative Prompt Tips

Using negative prompts reasonably can significantly improve generation quality:

Scenario 1: Portrait

{
 "negative_prompt": "blur, low quality, distortion, deformation, extra limbs, wrong proportions, ugly"
}

Scenario 2: Landscape Photo

{
 "negative_prompt": "blur, noise, overexposure, underexposure, low resolution, watermark"
}

Scenario 3: Product Image

{
 "negative_prompt": "blur, heavy shadows, excessive reflection, color distortion, messy background"
}

Scenario 4: Text Rendering

{
 "negative_prompt": "blurred text, typos, distorted font, unclear strokes"
}

5.4 Parameter Combination Suggestions

Usage Scenario	size	prompt_extend	watermark	negative_prompt
Rapid Creation	1328×1328	true	true	""
Precise Control	As needed	false	false	Detailed description
Text Rendering	1328×1328	false	true	"blurred text"
High Quality Output	1664×928	false	false	"blur, low quality, noise"
Poster Design	928×1664	true	false	"messy, low quality"

6. Billing Instructions

6.1 Billing Method

qwen-image-plus uses a per-image billing model:

Official Price: ¥0.20/image

System Price: ¥0.18/image (10% off)

Billing Unit: Billed based on the number of successfully generated images

Unified Price: Same price for all sizes

6.2 Billing Example

Generate 1 image: 1 image × ¥0.18 = ¥0.18
Generate 5 images: 5 images × ¥0.18 = ¥0.90

6.3 Billing Log Format

After successful generation, the system records billing info:

{
 "content": "Image generation: 1 image × ¥0.18 = ¥0.18",
 "modelName": "qwen-image-plus",
 "quota": 12329,
 "quotaDollar": "0.024658"
}

7. Frequently Asked Questions (FAQ)

Q1: Does qwen-image-plus support Image-to-Image?

A: No. qwen-image-plus is a pure Text-to-Image model. It only supports generating images from text descriptions and does not support uploading reference images. If you need Image-to-Image functionality, please use gemini-2.5-flash-image or doubao-seedream-4-0-250828.

Q2: How to generate images with Chinese text?

A: qwen-image-plus excels at Chinese/English text rendering. Simply state the text content clearly in the prompt:

{
 "prompt": "A poster, title is 'Double 11 Sale', subtitle '50% off site-wide', background is a red gradient"
}

The model will automatically render this text in the generated image.

Q3: How long is the generated image valid?

A: The image URL is valid for approximately 24 hours. It is recommended to download and save it immediately after receiving the response, or upload it to your own storage service.

Q4: Can I generate multiple images at once?

A: The current version of qwen-image-plus generates 1 image per request. If you need to generate multiple images, please initiate multiple concurrent requests.

Q5: Can prompt extension and negative prompts be used together?

A: Yes! They do not conflict:

prompt_extend: true - System expands your positive prompt

negative_prompt: "..." - Simultaneously applies negative constraints

Recommendation: Short prompt + Enable extension + Add negative prompt for best results.

Q6: How to ensure the generated image has no watermark?

A: Set watermark: false:

{
 "parameters": {
  "watermark": false
 }
}

Note: Removing watermarks may have usage restrictions according to Alibaba Cloud's terms of service; please comply with relevant regulations.

Q7: What image output formats are supported?

A: The image URL returned by qwen-image-plus points to a PNG format image, ensuring high-quality output.

Q8: How long does image generation take?

A: Usually completed within 5-15 seconds, depending on:

Prompt complexity

Image size

Server load

8. Full Request Examples

Example 1: Product Poster with Text

Example 2: Chinese Style Calligraphy Couplet

Example 3: Short Prompt + Auto Extension

The system will automatically extend "Cyberpunk city" into a detailed description to improve the generation result.

9. Performance Optimization Suggestions

9.1 Batch Generation

If you need to generate multiple images, concurrent requests are recommended to improve efficiency:

9.2 Prompt Optimization Strategy

Short Prompt: Enable prompt_extend: true to let the system optimize automatically.

Detailed Prompt: Disable prompt_extend: false to keep the original description.

Text Rendering: Explicitly mark text content with quotes in the prompt.

Add Negative Words: Exclude common defects (blur, low quality, etc.).

9.3 Caching Strategy

For requests with identical parameters, it is recommended to cache on the client side:

Cache key: ${prompt}_${size}_${negative_prompt}_${prompt_extend}

Cache time: 1-24 hours depending on business needs.

10. Quick Reference

Model Parameters Cheat Sheet

Parameter	Value
Model Name	qwen-image-plus
Supported Function	Text-to-Image (Pure text generation)
Core Advantage	Complex Text Rendering (Chinese/English)
Supported Sizes	1328×1328, 1664×928, 928×1664, 1472×1140, 1140×1472
Image Input	❌ Not Supported
Response Format	URL (Valid for approx. 24 hours)
Billing Method	Per image billing (¥0.18/image)

Core Functions Cheat Sheet

Function	Supported	Description
Text-to-Image	✅	Generate images from pure text descriptions
Text Rendering	✅	Excels at Chinese/English text rendering
Prompt Extension	✅	Optional intelligent extension
Negative Prompt	✅	Exclude unwanted elements
Watermark Control	✅	Select whether to add watermark
Image-to-Image	❌	Not Supported
Multi-Image Fusion	❌	Not Supported
Conversation Context	❌	Not Supported

Common Sizes Cheat Sheet

Size	Ratio	Scenario
1328×1328	1:1	Social media, avatars
1664×928	16:9	Video cover, landscape
928×1664	9:16	Short video, phone wallpaper
1472×1140	4:3	Landscape photo
1140×1472	3:4	Vertical poster

Parameter Combination Cheat Sheet

Scenario	size	prompt_extend	watermark	negative_prompt
Rapid Creation	1328×1328	true	true	""
Text Poster	928×1664	false	false	"blurred text"
High Quality Photo	1664×928	false	false	"blur, low quality, noise"
Couplet Calligraphy	1140×1472	false	true	"blurred text, typos"

Document Version: v1.0
Update Time: 2025-11-06
Model: Qwen Image Generation Enhanced Version (qwen-image-plus)
Technical Support: https://llm.ai-nebula.com

Appendix: Complete Request Example (Python)

Appendix: Complete Request Example (Node.js)

Technical Support

If you have any questions, please contact:

Documentation: https://llm.ai-nebula.com/docs

Technical Support: support@ai-nebula.com

Tongyi Qianwen Text to Image Model Interface Document

1. Interface Basic Information#

1.1 Core Parameters Description#

1.2 Request Parameter Structure#

Format 1: Simplified Format (Recommended)#

Format 2: Full Format#

1.3 Supported Image Sizes#

2. Text-to-Image Functionality#

2.1 Basic Text-to-Image (Square 1:1)#

2.2 Widescreen Image Generation (16:9)#

2.3 Vertical Image Generation (9:16)#

3. Advanced Features#

3.1 Text Rendering (Core Advantage)#

3.2 Negative Prompt#

3.3 Prompt Extension#

Enable Prompt Extension (Recommended)#

Disable Prompt Extension#

3.4 Watermark Control#

4. Response Handling#

4.1 Response Format#

4.2 Response Field Description#

4.3 Error Handling#

5. Best Practices#

5.1 Size Selection Suggestions#

5.2 Prompt Optimization Suggestions#

Text Rendering Prompt Tips#

General Prompt Tips#

5.3 Negative Prompt Tips#

Scenario 1: Portrait#

Scenario 2: Landscape Photo#

Scenario 3: Product Image#

Scenario 4: Text Rendering#

5.4 Parameter Combination Suggestions#

6. Billing Instructions#

6.1 Billing Method#

6.2 Billing Example#

6.3 Billing Log Format#

7. Frequently Asked Questions (FAQ)#

Q1: Does qwen-image-plus support Image-to-Image?#

Q2: How to generate images with Chinese text?#

Q3: How long is the generated image valid?#

Q4: Can I generate multiple images at once?#

Q5: Can prompt extension and negative prompts be used together?#

Q6: How to ensure the generated image has no watermark?#

Q7: What image output formats are supported?#

Q8: How long does image generation take?#

8. Full Request Examples#

Example 1: Product Poster with Text#

Example 2: Chinese Style Calligraphy Couplet#

Example 3: Short Prompt + Auto Extension#

9. Performance Optimization Suggestions#

9.1 Batch Generation#

9.2 Prompt Optimization Strategy#

9.3 Caching Strategy#

10. Quick Reference#

Model Parameters Cheat Sheet#

Core Functions Cheat Sheet#

Common Sizes Cheat Sheet#

Parameter Combination Cheat Sheet#

Appendix: Complete Request Example (Python)#

Appendix: Complete Request Example (Node.js)#

Technical Support#