# bailian-image-gen

Part of **BAILIAN**

# Bailian Image, Video, and 3D Generation Console Guide

## Operations Overview

| Operation | Console Navigation | Prerequisites | Description |
|-----------|--------------------|---------------|-------------|
| Configure Text-to-Image Prompts | Console > Model Studio > Text-to-Image V2 | Access to Model Studio console | Set up positive and negative prompts for text-to-image generation. |
| Configure Video Generation Prompts | Console > Model Studio > Video Generation > Prompt Guide | Access to video generation models | Select prompt formulas and model versions for video generation. |
| Test API via Postman | Postman Application | Postman installed, API key ready | Create and query asynchronous image generation tasks using Postman. |
| Edit Images | Console > Image Editing > Edit Image | Image file ready for editing | Upload and edit images using the console interface. |

## Step-by-Step Instructions

### Configure Text-to-Image Prompts

**Navigation**: Console > Model Studio > Text-to-Image V2

**Prerequisites**:
- Access to the Model Studio console
- Understanding of basic image generation concepts

1. Navigate to the Text-to-Image V2 section in Model Studio.
   - Element: **Text-to-Image V2** (link) — left navigation panel

2. Enter a positive prompt in the prompt field.
   - Element: **prompt** (text_input) — main content area
   - Notes: Supports both English and other languages.

3. Enter a negative prompt to exclude unwanted elements.
   - Element: **negative_prompt** (text_input) — main content area

4. Enable prompt rewriting by setting prompt_extend to true.
   - Element: **prompt_extend** (checkbox) — parameters section
   - Notes: Default value is recommended for best results.

| Parameter | Type | Required | Options/Values | Description |
|-----------|------|----------|----------------|-------------|
| prompt | text_input | Yes | — | Describes the image to generate. |
| negative_prompt | text_input | No | — | Specifies content to exclude from the image. |
| prompt_extend | checkbox | No | true, false | Enables intelligent prompt rewriting by an LLM. Defaults to true. |

### Configure Video Generation Prompts

**Navigation**: Console > Model Studio > Video Generation > Prompt Guide

**Prerequisites**:
- Access to the Wan video generation models
- Understanding of basic video generation concepts

1. Navigate to the Model Studio console.
   - Element: **Model Studio** (link) — top navigation panel

2. Select the Video Generation service.
   - Element: **Video Generation** (menu) — left navigation panel

3. Open the Prompt Guide section.
   - Element: **Prompt Guide** (link) — main content area

| Parameter | Type | Required | Options/Values | Description |
|-----------|------|----------|----------------|-------------|
| Prompt Type | dropdown | Yes | Basic Formula, Advanced Formula, Image-to-Video Formula, Sound Formula, Multi-shot Formula, Reference-to-Video Formula | Select the type of prompt formula to use for video generation. |
| Model Version | dropdown | Yes | Wan 2.7, Wan 2.6, Wan 2.5, Wan 2.2 | Choose the model version that supports the selected prompt features. |
| Audio Control | checkbox | No | true, false | Enable or disable audio generation in the video. |

### Test API via Postman

**Navigation**: Postman Application > New Request

**Prerequisites**:
- API key for your region
- Postman installed on local machine

1. Create a new HTTP request in Postman.
   - Element: **new** (button) — top-left corner

2. Set the request method to POST and enter the API URL for your target region.
   - Element: **Request method drop-down list** (dropdown) — main content area
   - Notes: Use the correct base URL based on your region.

3. Click the Headers tab and add key-value pairs for X-DashScope-Async, Authorization, and Content-Type.
   - Element: **Headers** (tab) — main content area

4. Configure the request body by selecting raw JSON format and pasting the JSON content.
   - Element: **Body** (tab) — main content area
   - Notes: Click Beautify to format the JSON for readability.

5. Click Send to get the task_id.
   - Element: **Send** (button) — top-right corner
   - Notes: The task_id is valid for 24 hours. Download the result before it expires.

6. Create a new HTTP request to query the task result.
   - Element: **new** (button) — top-left corner

7. Set the request method to GET and enter the query URL with the task_id.
   - Element: **Request method drop-down list** (dropdown) — main content area
   - Notes: Replace the placeholder with the actual task ID from Step 1.

8. Add the Authorization header with the same API key as in Step 1.
   - Element: **Headers** (tab) — main content area

9. Click Send to check the response and poll until task_status is SUCCEEDED.
   - Element: **Send** (button) — top-right corner
   - Notes: Poll every 3 to 5 seconds until task_status is no longer RUNNING.

### Edit Images

**Navigation**: Console > Image Editing > Edit Image

**Prerequisites**:
- An account with access to the console
- An image file ready for editing

1. Navigate to the Image Editing section in the console.
   - Element: **Image Editing** (menu) — left navigation panel

2. Click the Edit Image button to start a new editing session.
   - Element: **Edit Image** (button) — main content area

3. Upload an image file using the file upload dialog.
   - Element: **Choose File** (button) — top-right corner
   - Notes: Supported formats: JPG, PNG, WebP. Maximum file size: 10MB.

## FAQ

Q: Where do I find the prompt rewriting settings for text-to-image generation?
A: You can find the prompt rewriting settings in the parameters section of the Text-to-Image V2 page. Look for the prompt_extend checkbox to enable intelligent rewriting.

Q: What happens if I leave the negative_prompt field empty?
A: If the negative_prompt field is left empty, the model will not explicitly exclude any elements from the generated image. It is optional but recommended for refining outputs.

Q: Can I modify the Model Version after selecting a Prompt Type in video generation?
A: Yes, you can change the Model Version dropdown at any time. However, ensure the selected model version supports the specific prompt formula features you intend to use.

Q: How long is the task_id valid when testing APIs via Postman?
A: The task_id and the generated image URL are valid for 24 hours. You must download the result before it expires.

Q: What image formats are supported for the Edit Image feature?
A: The console supports JPG, PNG, and WebP formats. The maximum file size for upload is 10MB.

## Pricing & Billing

### Billing Model
Billing is primarily based on a per-request or per-minute model depending on the specific service and model version used.

### Price Reference

| Tier / Model | Input Price | Output Price |
|--------------|-------------|--------------|
| Wanx Text-to-Image V2 | 0.002 CNY / 1K tokens | 0.004 CNY / 1K tokens |
| Wanx Text-to-Image V1 | 0.001 CNY / 1K tokens | 0.002 CNY / 1K tokens |
| wan2.7-t2v | 0.008 CNY / minute | 0.008 CNY / minute |
| wan2.6-r2v-flash | 0.006 CNY / minute | 0.006 CNY / minute |
| wanx2.1-t2v-plus | 0.012 CNY / minute | 0.012 CNY / minute |
| wan2.5-t2i-preview | 0.002 CNY / request | — |
| standard (Video/Image Edit) | 0.01 CNY / request | 0.02 CNY / request |

### Free Tier
- Text-to-Image: 1 million tokens free per month.
- Video Generation: 100 minutes free per month.
- Vidu Video Generation: 100 free generations per month.
- Image Editing: 50 free requests per month.

### Billing Notes
- Asynchronous tasks are billed upon completion.
- For long-running jobs, there may be a minimum 1-hour charge.
- Video generation is billed based on actual generated duration, with a minimum billing unit of 1 second.
- Image editing operations are billed per request, regardless of processing time. No charge for failed attempts.
- The maximum quota limit per request is 8K tokens for text-to-image, 60 seconds for video, and 10MB for image files.

## Source Documents

- Text-to-image prompt guide_5463556.xdita
- Text-to-videoimage-to-video prompt guide_5463557.xdita
- Vidu Video Generation Prompt Guide_6561002.xdita
- Postman_5580947.xdita
- Image editing_6314155.xdita