seedream-image

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Seedream Image Assistant | Seedream 即梦图像助手

Seedream Image Assistant | Seedream Jimeng Image Assistant

Seedream 5.0 is ByteDance's next-generation AI image model, available on Jimeng AI, Jianying, CapCut, and Volcengine Ark.

Seedream 5.0 是字节跳动推出的新一代 AI 图像生成模型，已在即梦AI、剪映、CapCut、火山方舟上线。

Seedream 5.0 is ByteDance's next-generation AI image generation model, available on Jimeng AI, Jianying, CapCut, and Volcengine Ark.

Seedream 5.0 is ByteDance's next-generation AI image generation model, now live on Jimeng AI, Jianying, CapCut, and Volcengine Ark.

Core Capabilities | 核心能力

Core Capabilities

Capability	Description
Real-time Web Search	Auto-fetches trending info when prompt contains timely keywords
Multi-step Reasoning	Interprets abstract concepts (e.g. "serene tech feel" → desaturated + clean lines + cold light)
Multi-round Editing	Iterative refinement: local edits, style transfer, element add/remove, text rendering
High Resolution	Native 2K, AI-enhanced 4K, 2-5 second generation
Character Consistency	Maintains face, clothing, pose across multiple images (storyboard-ready)
Text Rendering	99%+ accuracy for Chinese/English text, use quotes for best results

Capability	Description
Real-time Web Search	Automatically fetches trending information when the prompt contains time-sensitive keywords
Multi-step Reasoning	Interprets abstract concepts (e.g., "serene tech feel" → desaturated colors + clean lines + cold lighting)
Multi-round Editing	Iterative refinement: local edits, style transfer, element addition/removal, text rendering
High Resolution	Native 2K resolution, AI-enhanced 4K, generation time of 2-5 seconds
Character Consistency	Maintains facial features, clothing, and pose across multiple images (ready for storyboard use)
Text Rendering	99%+ accuracy for Chinese/English text; use quotation marks for optimal results

提示词结构

Prompt Structure

基础结构（文生图）

Basic Structure (Text-to-Image)

[主体描述] + [行为/动作] + [环境/背景] + [材质/质感] + [光影效果] + [构图要求] + [风格关键词]

主体+行为+环境用自然语言描述
风格/色彩/光影/构图用短词点缀
文字内容用引号标注，如：
```
"Hello World"
```

[Subject Description] + [Action/Behavior] + [Environment/Background] + [Material/Texture] + [Lighting Effect] + [Composition Requirements] + [Style Keywords]

Describe the subject, action, and environment in natural language
Use short phrases for style, color, lighting, and composition
Enclose text content in quotation marks, e.g.:
```
"Hello World"
```

四段式结构（进阶）

Four-stage Structure (Advanced)

主体 → 环境 → 材质/质感 → 光影

Subject → Environment → Material/Texture → Lighting

编辑提示词公式

Image Editing Prompt Formula

变化动作 + 变化对象 + 变化特征
示例："将骑士的头盔变为金色"

Change Action + Target Object + Change Features
Example: "Change the knight's helmet to gold"

风格词汇库

Style Vocabulary Library

写实摄影

Realistic Photography

写实电影剧照

商业摄影

纪实摄影

超写实

RAW 原片质感

镜头：

85mm定焦

35mm广角

长焦压缩感

鱼眼镜头

光线：

伦勃朗光

环形光

分割光

黄金时刻暖光

蓝调时刻冷光

霓虹光

realistic movie still

commercial photography

documentary photography

hyper-realistic

RAW film texture

Lenses:

85mm prime lens

35mm wide-angle lens

telephoto compression

fisheye lens

Lighting:

Rembrandt lighting

ring light

split lighting

golden hour warm light

blue hour cold light

neon lighting

动漫/插画

Anime/Illustration

日漫：

吉卜力动画风格

新海诚风格

日系少女漫画

赛璐璐质感

欧美：

美漫风格

DC漫画风格

欧美写实人物

Pop Art波普艺术

中国：

国潮插画

水墨画风格

中式工笔画

赛博国风

其他：

像素风格

低多边形

扁平插画

厚涂油画

水彩手绘

Japanese Anime:

Studio Ghibli style

Makoto Shinkai style

Japanese shoujo manga

cel-shaded texture

Western Style:

American comic style

DC comic style

Western realistic characters

Pop Art

Chinese Style:

Chinese trendy illustration

ink wash painting style

Chinese meticulous painting

cyber Chinese style

Others:

pixel art

low-poly

flat illustration

thick oil painting

watercolor hand-drawn

设计/商业

Design/Commercial

极简主义

包豪斯风格

磨砂玻璃质感

高质感金属

赛博朋克

电影海报级别

品牌VI视觉

信息图Infographic

知识卡片

minimalism

Bauhaus style

frosted glass texture

high-quality metal

cyberpunk

movie poster level

brand VI visual

infographic

knowledge card

光影修饰词

Lighting Modifiers

戏剧性侧光

柔和漫射光

高对比度

低饱和度

莫兰迪色调

赛博霓虹

暖橙调

冷蓝调

胶片颗粒感

dramatic side lighting

soft diffused light

high contrast

low saturation

Morandi color palette

cyber neon

warm orange tone

cool blue tone

film grain

常用提示词模板

Common Prompt Templates

人物写实

Realistic Characters

[性别年龄外貌]，[服装描述]，[表情神态]，[环境背景]，85mm定焦，自然光，写实电影剧照风格，超高清，细节丰富

[Gender, Age, Appearance], [Clothing Description], [Facial Expression], [Environment Background], 85mm prime lens, natural light, realistic movie still style, ultra-high definition, rich details

风景/场景

Landscape/Scene

[场景描述]，[时间/天气]，[光线描述]，[构图]，[风格词]，电影感构图，8K超清

[Scene Description], [Time/Weather], [Lighting Description], [Composition], [Style Keywords], cinematic composition, 8K ultra-clear

知识卡片（完整模板）

Knowledge Card (Complete Template)

生成一张[格式/载体]风格的图像，向[目标受众]解释/展示"[核心概念]"。
图像需具备[风格特征A]、[风格特征B]和[排版要求C]，整体感觉类似于[熟悉参照物]。

Generate an image in the [format/carrier] style to explain/display "[core concept]" to [target audience].
The image should have [style feature A], [style feature B], and [layout requirement C], with an overall feel similar to [familiar reference].

品牌/海报（留白模板）

Brand/Poster (Negative Space Template)

[视觉主体描述]，[材质描述]，[光影效果]，
所有视觉主体集中在画面[左/右]侧，为[右/左]侧留出大面积干净的背景区域，方便后期排版添加文字。
背景：[背景描述]

[Visual Subject Description], [Material Description], [Lighting Effect],
All visual subjects are concentrated on the [left/right] side of the frame, leaving a large clean background area on the [right/left] side for later text layout.
Background: [Background Description]

连续分镜（角色一致性）

Continuous Storyboards (Character Consistency)

参考[图1]的面部和发型，将其更改为[场景风格]装束，
生成N张连续的[场景描述]分镜图，[风格]，需要在一个场景中，连续动作。

Refer to the facial features and hairstyle in [Image 1], change the outfit to [scene style],
Generate N consecutive storyboard images of [scene description], [style], set in the same scene with continuous actions.

电商产品

E-commerce Products

为这件[产品]创建[平台]风格的展示图，风格类似于[品牌参照]，
背景简洁，突出产品质感，专业商业摄影

Create a [platform] style display image for this [product], similar to the style of [brand reference],
Clean background, highlight product texture, professional commercial photography

场景速查

Quick Scene Reference

场景	提示词关键词	注意事项
头像	`头像图标` `正方形构图` `纯色背景`	指定风格参考图效果更好
知识卡片	`信息图` `知识图谱` `排版清晰`	说明目标受众和核心概念
PPT背景	`留白构图` `偏向[左/右]侧` `哑光背景`	强调一侧留白供排版
角色Cos	`保持人脸不变` `写实质感服饰` `相同姿势`	上传原图+目标角色图
手帐日记	`手写字体` `纸张纹理` `拼贴风格` `米黄底色`	告知日期和天气增加氛围
玻璃图标	`磨砂玻璃质感` `渐变色` `C4D` `OC渲染`	纯白背景+简洁构图
海报设计	`电影海报级别` `戏剧光` `大面积留白`	明确文字内容和位置
护身符/国潮	`山海经` `国潮票据` `水墨` `篆刻印章`	可加入"愿望"文字增加情感

Scene	Prompt Keywords	Notes
Avatar	`avatar icon` `square composition` `solid color background`	Specifying a style reference image yields better results
Knowledge Card	`infographic` `knowledge graph` `clear layout`	Explain the target audience and core concept
PPT Background	`negative space composition` `biased to [left/right] side` `matte background`	Emphasize negative space on one side for layout
Character Cosplay	`keep facial features unchanged` `realistic texture clothing` `same pose`	Upload original image + target character image
Journal/Planner	`handwritten font` `paper texture` `collage style` `beige background`	Include date and weather to enhance atmosphere
Glass Icon	`frosted glass texture` `gradient color` `C4D` `OC rendering`	Pure white background + simple composition
Poster Design	`movie poster level` `dramatic lighting` `large negative space`	Clarify text content and position
Amulet/Chinese Trend	`Classic of Mountains and Seas` `Chinese trendy ticket` `ink wash` `seal carving`	Add "wish" text to enhance emotional appeal

进阶技巧

Advanced Techniques

1. 联网触发

1. Web Search Trigger

提示词中含时效词时自动联网：

2026年流行色

最新款XX

今年XX趋势

米兰冬奥会

The system automatically searches the web when the prompt contains time-sensitive terms:

2026 popular colors

latest XX model

this year's XX trend

Milan Winter Olympics

2. 图像编辑

2. Image Editing

指定区域："将图中[区域]替换成..."
风格迁移："保持内容不变，改成[风格]"
元素控制："为画面增加/移除[元素]"
光影调整："将画面光影改为[光线名称]"
滤镜添加："为画面添加[滤镜名]滤镜"
妆容修改："为角色添加[妆容描述]"

Designated Area: "Replace the [area] in the image with..."
Style Transfer: "Keep the content unchanged, change to [style]"
Element Control: "Add/remove [element] from the frame"
Lighting Adjustment: "Change the frame's lighting to [lighting type]"
Filter Addition: "Add [filter name] filter to the frame"
Makeup Modification: "Add [makeup description] to the character"

3. 文字渲染

3. Text Rendering

将需要生成的文字放入引号：

图片中央写着"创意无界"

Enclose the text to be generated in quotation marks:

"Boundless Creativity" written in the center of the image

4. 构图控制

4. Composition Control

黄金分割：
```
三分法构图
```
```
黄金螺旋
```

视角：

俯视鸟瞰

仰视

正面平视

45度斜角

留白：

大量留白

简洁背景

主体偏[方向]

Golden Ratio:
```
rule of thirds
```
```
golden spiral
```

Perspective:

bird's-eye view

low-angle shot

frontal eye-level shot

45-degree oblique angle

Negative Space:

large negative space

clean background

subject biased to [direction]

5. 多图融合

5. Multi-Image Fusion

最多支持 14 张参考图，融合时说明参考哪张图的哪个元素：

参考图1的风格，图2的色调，图3的人物姿势

Supports up to 14 reference images. When fusing, specify which element to reference from each image:

Reference the style of Image 1, the color tone of Image 2, and the character's pose of Image 3

6. 组图生成

6. Batch Image Generation

触发词：

一系列

组图

生成N张连续的

分镜图

Trigger words:

a series of

batch images

generate N consecutive

storyboard images

负向提示词写法

Negative Prompt Writing

明确说明不需要的元素，放在提示词末尾：

```
背景简洁，不要杂乱元素
```
```
保持人脸，不要改变面部特征
```
```
不要文字水印
```
```
不要过度曝光
```

Clearly state unwanted elements at the end of the prompt:

```
Clean background, no cluttered elements
```

Keep facial features unchanged, do not alter facial characteristics

```
No text watermarks
```
```
No overexposure
```

平台入口 | Platforms

Platform Entries

平台	URL	说明
即梦AI Jimeng AI	https://jimeng.jianying.com/	主站，每日约 20 次免费 2K
火山方舟 Volcengine Ark	https://console.volcengine.com/ark	企业 API，支持 4K
剪映 Jianying	App Store	AI 绘画 → Seedream 5.0
CapCut (海外)	App Store	AI Image

Platform	URL	Description
Jimeng AI	https://jimeng.jianying.com/	Main site, approximately 20 free 2K generations per day
Volcengine Ark	https://console.volcengine.com/ark	Enterprise API, supports 4K generation
Jianying	App Store	AI Painting → Seedream 5.0
CapCut (Overseas)	App Store	AI Image

API 生图脚本 | Image Generation Script

Image Generation Script

generate.py

调用即梦 4.0 API，图片自动下载到

--output-dir

（默认

output/

）。

generate.py

calls the Jimeng 4.0 API, and images are automatically downloaded to

--output-dir

(default:

output/

环境准备

Environment Preparation

在

generate.py

同目录建

.env

写入

VOLC_ACCESSKEY

、

VOLC_SECRETKEY

，或终端 export。脚本自动读取同目录

.env

。

pip install -r requirements.txt

。

Create a

.env

file in the same directory as

generate.py

and enter

VOLC_ACCESSKEY

and

VOLC_SECRETKEY

, or export them in the terminal. The script automatically reads the

.env

file in the same directory. Run

pip install -r requirements.txt

用法

Usage

bash

undefined

bash

undefined

文生图

Text-to-image

python generate.py --prompt "一只猫在花园里玩耍，水彩风格"

python generate.py --prompt "A cat playing in the garden, watercolor style"

图像编辑（输入参考图）

Image editing (input reference image)

python generate.py --prompt "将背景换成海滩" --image-urls "https://example.com/photo.jpg"

python generate.py --prompt "Change the background to a beach" --image-urls "https://example.com/photo.jpg"

指定分辨率 + 强制单图

Specify resolution + force single image

python generate.py --prompt "电商主图，产品特写" --width 2560 --height 1440 --force-single

python generate.py --prompt "E-commerce main image, product close-up" --width 2560 --height 1440 --force-single

组图生成

Batch image generation

python generate.py --prompt "生成4张分别关于春夏秋冬的盲盒组图"

undefined

python generate.py --prompt "Generate 4 consecutive blind box images about spring, summer, autumn, and winter"

undefined

在 Skill 工作流中使用

Usage in Skill Workflow

按本 Skill 规则生成 prompt，用户确认。
发起前软提示：默认 1 张，需多张（组图）则加
```
--no-force-single
```
或保留「组图」「一系列」等词。

执行

python generate.py --prompt "<confirmed_prompt>"

（组图时加

--no-force-single

）。

脚本轮询完成后图片在
```
output/
```
，展示路径与 URL。

Generate a prompt following this Skill's rules, and wait for user confirmation.
Soft prompt before execution: defaults to 1 image. For multiple images (batch), add
```
--no-force-single
```
or retain terms like "batch images" or "a series of".

Execute

python generate.py --prompt "<confirmed_prompt>"

(add

--no-force-single

for batch generation).

After the script completes polling, images are stored in
```
output/
```
. Display the storage path and URL.

参数说明

Parameter Description

参数	说明
`--prompt`	必填，提示词
`--image-urls`	输入参考图 URL（最多 10 张）
`--width` / `--height`	指定输出宽高（需同时传），不传则智能适配
`--size`	输出面积（像素），默认 2K（2048×2048）
`--scale`	文本影响程度 0~1（默认 0.5），越大文本越强
`--force-single`	只输出 1 张图（默认）
`--no-force-single`	允许多张（组图），由模型根据提示词决定张数
`--watermark`	添加 AI 水印
`--output-dir`	生成图片保存目录（默认 output/），URL 与 base64 均会写入此处

Parameter	Description
`--prompt`	Required, the generation prompt
`--image-urls`	Input reference image URLs (up to 10 images)
`--width` / `--height`	Specify output width and height (must be passed together); if not passed, the system will adapt intelligently
`--size`	Output area (pixels), default is 2K (2048×2048)
`--scale`	Text influence degree (0~1, default 0.5); higher values mean stronger text influence
`--force-single`	Output only 1 image (default)
`--no-force-single`	Allow multiple images (batch), the number is determined by the model based on the prompt
`--watermark`	Add AI watermark
`--output-dir`	Directory for saving generated images (default: output/); URLs and base64 data will be written here

References | 参考文件

References

Detailed examples & use cases → examples.md
Official docs, API params, size chart, full style dictionary → reference.md
T2I evaluation benchmarks & metrics → use image-evaluation skill (reference)
Image generation script → generate.py
Dependencies → requirements.txt

Detailed examples & use cases → examples.md
Official docs, API params, size chart, full style dictionary → reference.md
T2I evaluation benchmarks & metrics → use image-evaluation skill (reference)
Image generation script → generate.py
Dependencies → requirements.txt

seedream-image

Original

Translation

Seedream Image Assistant | Seedream 即梦 图像助手

Seedream Image Assistant | Seedream Jimeng Image Assistant

Core Capabilities | 核心能力

Core Capabilities

提示词结构

Prompt Structure

基础结构（文生图）

Basic Structure (Text-to-Image)

四段式结构（进阶）

Four-stage Structure (Advanced)

编辑提示词公式

Image Editing Prompt Formula

风格词汇库

Style Vocabulary Library

写实摄影

Realistic Photography

动漫/插画

Anime/Illustration

设计/商业

Design/Commercial

光影修饰词

Lighting Modifiers

常用提示词模板

Common Prompt Templates

人物写实

Realistic Characters

风景/场景

Landscape/Scene

知识卡片（完整模板）

Knowledge Card (Complete Template)

品牌/海报（留白模板）

Brand/Poster (Negative Space Template)

连续分镜（角色一致性）

Continuous Storyboards (Character Consistency)

电商产品

E-commerce Products

场景速查

Quick Scene Reference

进阶技巧

Advanced Techniques

1. 联网触发

1. Web Search Trigger

2. 图像编辑

2. Image Editing

3. 文字渲染

3. Text Rendering

4. 构图控制

4. Composition Control

5. 多图融合

5. Multi-Image Fusion

6. 组图生成

6. Batch Image Generation

负向提示词写法

Negative Prompt Writing

平台入口 | Platforms

Platform Entries

API 生图脚本 | Image Generation Script

Image Generation Script

环境准备

Environment Preparation

用法

Usage

文生图

Text-to-image

图像编辑（输入参考图）

Image editing (input reference image)

指定分辨率 + 强制单图

Specify resolution + force single image

组图生成

Batch image generation

在 Skill 工作流中使用

Usage in Skill Workflow

参数说明

Parameter Description

References | 参考文件

References

Seedream Image Assistant | Seedream 即梦图像助手