SJinn
Seedream V4.5

Seedream V4.5

Generate and edit high-quality images using ByteDance's Seedream V4.5 model

Tool Overview

The Seedream V4.5 tool uses ByteDance's Seedream V4.5 model to generate high-quality images from text descriptions. It also supports image editing by providing reference images alongside text prompts. Seedream V4.5 is an upgraded version with improved detail rendering and prompt adherence.

Tool Identifier

seedream-v4-5-api

Parameters

Required Parameters

prompt

  • Type: string (required)
  • Description: Text description of the image content to guide AI image generation or editing
  • Validation: Must be a non-empty string
  • Example: "A futuristic cyberpunk cityscape at night with neon lights"

Optional Parameters

image_list

  • Type: string[] (optional)
  • Default: [] (empty array, no reference images)
  • Format: Array of complete HTTP/HTTPS URLs (e.g., ["https://cdn.sjinn.ai/uploads/image.jpg"])
  • Description: Reference images for image editing. When provided, the model will edit based on these images and the prompt

aspect_ratio

  • Type: string (optional)
  • Default: "auto"
  • Supported Values:
    • "auto" - Automatically select the optimal aspect ratio
    • "16:9" - Widescreen landscape
    • "9:16" - Portrait mobile
    • "1:1" - Square format
    • "4:3" - Standard landscape
    • "3:4" - Standard portrait
    • "3:2" - Classic landscape
    • "2:3" - Classic portrait
  • Note: If "auto" is selected or an invalid value is provided, the aspect ratio will be automatically determined

Pricing

  • Credits Consumed: 50 credits per task
  • Membership Requirement: None

Request Examples

Text-to-Image (No Reference)

curl -X POST https://sjinn.ai/api/un-api/create_tool_task \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "tool_type": "seedream-v4-5-api",
    "input": {
      "prompt": "A futuristic cyberpunk cityscape at night with neon lights",
      "aspect_ratio": "16:9"
    }
  }'

Image Editing with Reference

curl -X POST https://sjinn.ai/api/un-api/create_tool_task \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "tool_type": "seedream-v4-5-api",
    "input": {
      "prompt": "Transform into an oil painting style with warm tones",
      "image_list": ["https://cdn.sjinn.ai/uploads/photo.jpg"],
      "aspect_ratio": "1:1"
    }
  }'

Using JavaScript

const response = await fetch('https://sjinn.ai/api/un-api/create_tool_task', {
  method: 'POST',
  headers: {
    'Authorization': 'Bearer YOUR_API_KEY',
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    tool_type: 'seedream-v4-5-api',
    input: {
      prompt: 'A futuristic cyberpunk cityscape at night with neon lights',
      aspect_ratio: '16:9',
    },
  }),
});

const result = await response.json();
console.log('Task ID:', result.data.task_id);

Using Python

import requests

url = 'https://sjinn.ai/api/un-api/create_tool_task'
headers = {
    'Authorization': 'Bearer YOUR_API_KEY',
    'Content-Type': 'application/json'
}
data = {
    'tool_type': 'seedream-v4-5-api',
    'input': {
        'prompt': 'A futuristic cyberpunk cityscape at night with neon lights',
        'aspect_ratio': '16:9'
    }
}

response = requests.post(url, json=data, headers=headers)
result = response.json()
print('Task ID:', result['data']['task_id'])

Response Examples

Success Response

{
  "success": true,
  "errorMsg": "",
  "error_code": 0,
  "data": {
    "task_id": "550e8400-e29b-41d4-a716-446655440000"
  }
}

Error Response

{
  "success": false,
  "errorMsg": "prompt is required",
  "error_code": 400
}

Best Practices

  1. Prompt Optimization:

    • Use detailed, specific descriptions for better results
    • Include style, color, lighting, and composition details
    • Can specify art styles (e.g., "photorealistic", "anime style", "oil painting")

  2. Reference Image Usage:

    • Image editing: Provide the original image to edit
    • Style reference: Provide style example images
    • Multi-image fusion: Can provide multiple reference images (recommend no more than 3)

  3. Aspect Ratio Selection:

    • auto is suitable for letting AI automatically choose the best ratio
    • Social media: Instagram uses 1:1, Story uses 9:16
    • Web banners: Use 16:9

  4. Image URL Format:

    • All image URLs in image_list must be complete HTTP/HTTPS addresses
    • Ensure images are publicly accessible

  5. Generation Time: Typically 10-60 seconds to complete

Related Links