gpt-image-2

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

GPT Image 2

A single Python entrypoint that covers every GPT Image 2 route, with strict pre-flight validation of the model's size, aspect, and feature constraints.

一个统一的Python入口点，覆盖GPT Image 2的所有路由，并对模型的尺寸、宽高比和功能约束进行严格的预校验。

Workflow

工作流程

Open references/config.md to pick environment variables and defaults.
Open references/api-surface.md to choose between
```
generations
```
,
```
edits
```
, and
```
responses
```
.
Prefer
```
OPENAI_BASE_URL=https://api.openai.com/v1
```
unless the user asks for a different OpenAI-compatible endpoint.
Use
```
gpt-image-2
```
for
```
generations
```
and
```
edits
```
; use a text-capable Responses model such as
```
gpt-5.4
```
for
```
responses
```
.
Run
```
scripts/gpt_image.py
```
with one of the three subcommands.
Add
```
--dry-run
```
first when the payload shape is the main risk.
Add
```
--save-response <path>
```
when the raw JSON body or SSE event stream needs to be kept for debugging.

打开references/config.md选择环境变量和默认配置。
打开references/api-surface.md选择
```
generations
```
、
```
edits
```
或
```
responses
```
模式。
除非用户要求使用其他兼容OpenAI的端点，否则优先使用
```
OPENAI_BASE_URL=https://api.openai.com/v1
```
。
```
generations
```
和
```
edits
```
模式使用
```
gpt-image-2
```
；
```
responses
```
模式使用支持文本功能的Responses模型，例如
```
gpt-5.4
```
。
运行
```
scripts/gpt_image.py
```
并指定三个子命令之一。
当请求负载格式是主要风险时，先添加
```
--dry-run
```
参数。
当需要保存原始JSON响应体或SSE事件流用于调试时，添加
```
--save-response <path>
```
参数。

Commands

命令示例

Text-to-image through the public Images API:

powershell

python .\skills\gpt-image-2\scripts\gpt_image.py generations `
  --prompt "A bold product hero image for a developer tool homepage" `
  --output .\out\hero.png `
  --size 1536x1024 `
  --quality high `
  --format png

Multi-image batch with a filename pattern:

powershell

python .\skills\gpt-image-2\scripts\gpt_image.py generations `
  --prompt "A cinematic city skyline at night" `
  --output .\out\skyline-{index}.webp `
  --n 3 `
  --format webp `
  --compression 90

Image edits with two inputs plus a mask:

powershell

python .\skills\gpt-image-2\scripts\gpt_image.py edits `
  --prompt "Blend the two references into one clean marketing illustration" `
  --image .\refs\subject.png `
  --image .\refs\background.png `
  --mask .\refs\mask.png `
  --output .\out\edit-{index}.png `
  --image-field-style brackets `
  --n 2

Responses API with streaming and partial previews:

powershell

python .\skills\gpt-image-2\scripts\gpt_image.py responses `
  --input-text "Generate a poster for an AI developer summit" `
  --model gpt-5.4 `
  --output .\out\poster-{index}.png `
  --stream `
  --partial-images 2 `
  --save-response .\out\poster-events.json

Responses API edit with a local image plus a mask:

powershell

python .\skills\gpt-image-2\scripts\gpt_image.py responses `
  --input-text "Turn this product shot into a clean studio ad" `
  --model gpt-5.4 `
  --input-image .\refs\product.png `
  --mask .\refs\mask.png `
  --output .\out\studio.png `
  --action edit

Inspect the built request without sending it:

powershell

python .\skills\gpt-image-2\scripts\gpt_image.py generations `
  --prompt "A minimal cover image" `
  --output .\out\cover.png `
  --dry-run

通过公开图像API实现文本转图像：

powershell

python .\skills\gpt-image-2\scripts\gpt_image.py generations `
  --prompt "A bold product hero image for a developer tool homepage" `
  --output .\out\hero.png `
  --size 1536x1024 `
  --quality high `
  --format png

带文件名模板的多图像批量生成：

powershell

python .\skills\gpt-image-2\scripts\gpt_image.py generations `
  --prompt "A cinematic city skyline at night" `
  --output .\out\skyline-{index}.webp `
  --n 3 `
  --format webp `
  --compression 90

双输入加蒙版的图像编辑：

powershell

python .\skills\gpt-image-2\scripts\gpt_image.py edits `
  --prompt "Blend the two references into one clean marketing illustration" `
  --image .\refs\subject.png `
  --image .\refs\background.png `
  --mask .\refs\mask.png `
  --output .\out\edit-{index}.png `
  --image-field-style brackets `
  --n 2

支持流式传输和局部预览的Responses API：

powershell

python .\skills\gpt-image-2\scripts\gpt_image.py responses `
  --input-text "Generate a poster for an AI developer summit" `
  --model gpt-5.4 `
  --output .\out\poster-{index}.png `
  --stream `
  --partial-images 2 `
  --save-response .\out\poster-events.json

带本地图像和蒙版的Responses API编辑：

powershell

python .\skills\gpt-image-2\scripts\gpt_image.py responses `
  --input-text "Turn this product shot into a clean studio ad" `
  --model gpt-5.4 `
  --input-image .\refs\product.png `
  --mask .\refs\mask.png `
  --output .\out\studio.png `
  --action edit

检查构建好的请求而不发送：

powershell

python .\skills\gpt-image-2\scripts\gpt_image.py generations `
  --prompt "A minimal cover image" `
  --output .\out\cover.png `
  --dry-run

Rules

规则说明

Use
```
generations
```
for public text-to-image calls.
Use
```
edits
```
for multipart image edits and mask uploads.
Use
```
responses
```
for advanced flows: streaming, mixed text + image input,
```
previous_response_id
```
,
```
tool_choice
```
,
```
action
```
, and optional
```
tool_model
```
.
Process environment variables override
```
.env
```
; CLI flags override both.
Never print secrets.
```
--output
```
takes either a single path or a pattern such as
```
image-{index}.png
```
for multi-image or streaming flows.
```
responses
```
uses a top-level Responses model separate from the image model; default it to
```
gpt-5.4
```
unless you need another text-capable model.
```
quality
```
on Responses tool flows is passed through, but final behavior still depends on the hosted image tool.
On OpenAI GPT image models, omit
```
response_format
```
; image data already comes back as base64.
Fail fast on unsupported
```
gpt-image-2
```
combinations: transparent background, invalid size,
```
partial_images
```
outside
```
0..3
```
, or
```
stream=true
```
with
```
n>1
```
on public Images routes.

公开文本转图像调用使用
```
generations
```
模式。
多部分图像编辑和蒙版上传使用
```
edits
```
模式。
高级流程使用
```
responses
```
模式：流式传输、文本+图像混合输入、
```
previous_response_id
```
、
```
tool_choice
```
、
```
action
```
以及可选的
```
tool_model
```
。
进程环境变量优先级高于
```
.env
```
文件；CLI参数优先级高于两者。
绝不打印敏感信息。
```
--output
```
参数可接受单个路径或模板（如
```
image-{index}.png
```
），用于多图像或流式传输流程。
```
responses
```
模式使用独立于图像模型的顶层Responses模型；默认使用
```
gpt-5.4
```
，除非需要其他支持文本功能的模型。
Responses工具流程中的
```
quality
```
参数会直接传递，但最终行为仍取决于托管的图像工具。
在OpenAI GPT图像模型上，省略
```
response_format
```
；图像数据已以base64格式返回。
遇到不支持的
```
gpt-image-2
```
组合时立即终止：透明背景、无效尺寸、
```
partial_images
```
超出0..3范围，或公开图像路由中
```
stream=true
```
且
```
n>1
```
的情况。

Resources

资源链接

Script: scripts/gpt_image.py
Config reference: references/config.md
API surface reference: references/api-surface.md

脚本：scripts/gpt_image.py
配置参考：references/config.md
API接口参考：references/api-surface.md