seedream 5.0 lite

seedream 5.0 lite

Overview

Based on Volcano Ark seedream 5.0 lite model AI image generation tool. Through the redfox.hk platform, it encapsulates the complex ARK API authentication process,generate high-quality images with a single command。

Why use this Skill?

For developers: no need to apply for a whitelist on Volcano Ark, configure AK/SK, or integrate ARK authentication—one command does it all
For creators: no need to buy a membership or bind a card for a subscription; pay per use, leave when done

Target users

Designers, social media creators, content operations staff, AI enthusiasts, and any users who need to quickly turn ideas into images.

Features

Core features

Text-to-image: enter a Chinese or English description, and seedream 5.0 lite automatically generates a high-definition imageText-to-image: enter a Chinese or English description, and seedream 5.0 lite automatically generates a high-definition image
Image-to-image: upload a reference image + prompt to edit and transform the style based on the original imageImage-to-image: upload a reference image + prompt to edit and transform the style based on the original image
Multi-image generationMulti-image generation: supports auto mode to automatically generate multiple related images (up to 15)
Prompt optimizationPrompt optimization: built-in standard(high quality) and fast(fast) two optimization modes
High-resolution outputHigh-resolution output: supports 2K/3K/4K or custom pixel dimensions, default 2048x2048
Task managementTask management: supports obtaining a taskId after submitting a task, and you can query task progress and results at any time
Automatic pollingAutomatic polling: the script automatically polls task status (queued/generating/completed), no manual waiting required

Technical highlights

Underlying model:doubao-seedream-5-0-260128
Output formats: PNG, JPEG
Watermark control: choose whether to add the "AI-generated" watermark
Image URLs are automatically transferred to OSS for long-term validity
HTTPS secure transmission: full-chain SSL verification

Parameter quick reference

Parameter	Description	Default value
`prompt`	Image generation prompt (required, Chinese or English both acceptable)	—
`--image`	Reference image path or URL (enables image-to-image mode)	—
`--size`	Image size:`2K`/`3K`/`4K` or specific pixels such as `2048x2048`	`2048x2048`
`--format`	Output format:`png` / `jpeg`	`jpeg`
`--watermark`	Add the "AI-generated" watermark	Not added by default
`--sequential`	Multi-image mode:`auto` / `disabled`	`disabled`
`--max-images`	Maximum number of images to generate in multi-image mode (1-15)	`4`
`--optimize`	Prompt optimization:`standard` / `fast`	—
`-o, --output-dir`	Output directory	`~/Downloads/QoderImages`
`--prefix`	Filename prefix	`image`
`--no-download`	Submit only without waiting	—
`--task-id`	Query an existing task	—

Use cases

Use case 1: Social media content creation

Role: Xiaohongshu blogger / self-media operatorRole: Xiaohongshu blogger / self-media operator

Needs: quickly generate high-quality accompanying images, cover images, and product display imagesNeeds: quickly generate high-quality accompanying images, cover images, and product display images

How to use：

python3 "$SKILL_PATH/assets/seedream.py" "一杯拿铁咖啡放在木质桌面上，柔和的自然光，ins风" --size 2048x2048

Expected benefits：From finding an image/taking a photo to generating an image takes only a few seconds, and it supports batch group image generation

Scenario 2: E-commerce Product Display

Role：E-commerce operations / independent site seller

Needs：Generate product scene images with a unified style for products

How to use：

python3 "$SKILL_PATH/assets/seedream.py" "白色无线耳机悬浮在浅灰色背景上，专业产品摄影，柔和阴影" --format png --size 3K

Expected benefits：Zero photography cost, quickly produce high-quality product images

Scenario 3: Rapid Validation of Design Inspiration

Role：UI designer / illustrator

Needs：Quickly turn the image in your mind into a visual image using natural language, and validate the creative direction

How to use：

python3 "$SKILL_PATH/assets/seedream.py" "赛博朋克风格的未来城市，霓虹灯倒映在雨后的街道上，电影级构图" --size 4K

Expected benefits：Lower the creative threshold; from idea to visual draft requires only one line of command

Scenario 4: Group Image Content Operations

Role：Content operations / marketing personnel

Needs：Generate multiple images with a unified style at once for articles/tweets/event pages

How to use：

python3 "$SKILL_PATH/assets/seedream.py" "一组清新风格的早餐摄影，包含面包、水果、咖啡" --sequential auto --max-images 6

Expected benefits：Batch generate visually consistent visual assets and improve content production efficiency

Scenario 5: Image-to-Image Style Transfer

Role：Photographer / art creator

Needs：Perform style transfer or element modification based on existing images

How to use：

python3 "$SKILL_PATH/assets/seedream.py" "将画面转换成宫崎骏动画风格，色彩明亮温暖" --image ~/Pictures/photo.jpg

Expected benefits：No need to master complex image editing software; natural language can drive style transfer

Dependency Installation

Dependencies	Installation command
`requests`	`pip3 install requests`

First-time use

After configuring the API Key, you can use it:

# 设置环境变量
export REDFOX_API_KEY=ak_你的密钥

# 运行
python3 "$SKILL_PATH/assets/seedream.py" "一只橘色的猫咪坐在窗台上看着窗外的夕阳"

Go to redfox.hk to register and obtain an API Key.

Subsequent use

Go to redfox.hk to register an account and obtain your own API Token. Choose one of the three configuration methods:

Configuration method	Description	Command
Environment variable(Recommended)	Set once, effective globally	`export REDFOX_API_KEY=ak_你的密钥`
Command-line parameter	Temporary use, effective for a single time	`python3 "$SKILL_PATH/assets/seedream.py" "prompt" --api-key ak_你的密钥`
Configuration file	Persistent storage, retained across sessions	`mkdir -p ~/.qoder/apis && echo '{"api_key":"ak_你的密钥"}' > ~/.qoder/apis/redfox.json`

Usage Guide

Basic Usage

1. Enter a prompt to generate an image

python3 "$SKILL_PATH/assets/seedream.py" "一只橘猫在窗台上打哈欠，阳光温暖地照在它的毛上"

The script will automatically submit the task and poll while waiting (about 10-60 seconds). After completion, it will download to ~/Downloads/QoderImages/。

3. View the result

After successful generation, the image information and file path will be displayed:

[✓] Model: doubao-seedream-5-0-260128, Size: 2048x2048
[✓] Generated images: 1, Tokens: 1024
[→] Downloading 1/1: image.jpeg
[████████████████████] 100%

✓ Done!
  /Users/you/Downloads/QoderImages/image.jpeg (312.5 KB)

Advanced Usage

High-resolution output

# 4K 超高清
python3 "$SKILL_PATH/assets/seedream.py" "雪山日出，金色阳光洒在雪顶" --size 4K

# 指定具体像素
python3 "$SKILL_PATH/assets/seedream.py" "城市夜景" --size 2048x1152

Group image generation (batch image output)

# 自动生成最多 6 张关联图片
python3 "$SKILL_PATH/assets/seedream.py" "一组极简风办公桌面静物" --sequential auto --max-images 6

Image-to-image editing

# 基于参考图修改
python3 "$SKILL_PATH/assets/seedream.py" "把背景换成海边日落" --image ~/Pictures/portrait.jpg

# 使用网络图片作为参考
python3 "$SKILL_PATH/assets/seedream.py" "转换成油画风格" --image "https://example.com/photo.jpg"

Prompt optimization

# 标准模式（质量更高，耗时较长）
python3 "$SKILL_PATH/assets/seedream.py" "未来太空站内部" --optimize standard

# 快速模式（快速出图，质量一般）
python3 "$SKILL_PATH/assets/seedream.py" "快速草图概念" --optimize fast

Task management

# 仅提交任务，返回 taskId 后立即退出
python3 "$SKILL_PATH/assets/seedream.py" "复杂场景" --no-download

# 稍后用 taskId 查询结果
python3 "$SKILL_PATH/assets/seedream.py" "ignored" --task-id ark_abc123def456

Command Quick Reference

Command	Function
`python3 seedream.py "提示词"`	Basic text-to-image
`--size 4K`	4K high resolution
`--format png`	PNG format output
`--sequential auto --max-images 6`	Generate 6 images in group mode
`--image ~/pic.jpg`	Image-to-image mode
`--optimize standard`	Standard prompt optimization
`--watermark`	Add watermark
`--no-download`	Submit task only
`--task-id <id>`	Query an existing task
`--api-key <key>`	Specify API Key
`-o ~/Desktop`	Specify output directory

Project Architecture

Directory Structure

seedream-5-lite/
├── SKILL.md              # Skill 定义与文档
└── assets/               # 工具脚本
    └── seedream.py       # 图片生成主程序

Tech Stack

Components	Technology
Runtime Environment	Python 3.6+
HTTP Library	requests
API Platform	redfox.hk
Underlying Model	Volcano Ark seedream 5.0 lite (`doubao-seedream-5-0-260128`)
Output Format	PNG、JPEG

Core Modules

Module	Responsibility
`get_api_key()`	Obtain API Key with three-level priority: CLI > environment variables > configuration file
`upload_image()`	Upload the local reference image to OSS and obtain an accessible URL
`submit_task()`	Build the request body and submit the ARK image generation task
`poll_result()`	Poll task status (queued/running/succeeded/failed), up to 6 minutes
`download_images()`	Stream download the generated image with a progress bar display
`main()`	CLI entry: argument parsing, API Key validation, dual-branch (submit/query) flow

Data Flow

用户输入提示词 → submit_task() → redfox.hk API → 火山方舟 seedream 5.0 lite
                                                                      ↓
用户获得图片 ← download_images() ← poll_result() ← 任务完成回调

FAQ

Installation-Related Questions

Q1: What are the features of this Skill?

A: Call the seedream 5.0 lite model directly from the command line, supporting text-to-image, image-to-image, image set generation, prompt optimization, and high-resolution output.

Q2: How do I obtain an API Key?

A: Go to redfox.hk to register and obtain your own API Token.

Q3: Where is the configuration file placed?

A: Place it in ~/.qoder/apis/redfox.json, with the content format:{"api_key": "ak_你的密钥"}。

Q4: How do I verify whether the API Key is configured successfully?

A: Run python3 seedream.py "测试" --no-download. If it returns a taskId, the configuration is successful.

Usage-Related Questions

Q5: How long does it take to generate an image?

A: Usually 10-60 seconds; complex scenes or high resolution may take longer. The script will automatically poll and wait, up to 6 minutes.

Q6: Does it support uploading my own reference image?

A: Yes. Pass in the local image path via the --image parameter (such as --image ~/Pictures/photo.jpg). The script will automatically upload it to OSS and then submit the image-to-image task. Directly passing a web image URL is also supported.

Q7: Which prompt languages are supported?

A: Both Chinese and English are supported; the API handles this automatically internally.

Q8: Where are the output files saved?

A: By default, they are saved to ~/Downloads/QoderImages/image.jpeg, and the directory can be specified via the -o parameter.

Q9: What is the difference between image set mode and single image mode?

A：--sequential disabled(default) generates only a single image;--sequential auto will automatically determine based on the prompt and generate multiple related images, up to --max-images images.

Troubleshooting

Q10: What should I do if the task times out?

A: The script waits up to 6 minutes. After timeout, it will print the taskId, and you can later query and download the result again via the --task-id parameter.

Q11: It says "API request failed"?

A: Check whether the network connection is normal and confirm that the redfox.hk service is accessible. If it continues to fail, the API Key may have expired or there may be insufficient balance.

Q12: Image download failed?

A: Confirm that the output directory has write permission and that there is sufficient disk space. If the OSS link has expired, you can use --task-id to query again and obtain a new download link.

Getting Help

If you have other questions, you can go to redfox.hk View the platform documentation or contact customer service.

About this skill

Overview

Why use this Skill?

Target users

Features

Core features

Technical highlights

Parameter quick reference

Use cases

Use case 1: Social media content creation

Scenario 2: E-commerce Product Display

Scenario 3: Rapid Validation of Design Inspiration

Scenario 4: Group Image Content Operations

Scenario 5: Image-to-Image Style Transfer

Dependency Installation

First-time use

Subsequent use

Usage Guide

Basic Usage

1. Enter a prompt to generate an image

3. View the result

Advanced Usage

High-resolution output

Group image generation (batch image output)

Image-to-image editing

Prompt optimization

Task management

Command Quick Reference

Project Architecture

Directory Structure

Tech Stack

Core Modules

Data Flow

FAQ

Installation-Related Questions

Usage-Related Questions

Troubleshooting

Getting Help