Generate and edit images with GPT Image 2.
Generates and edits images through GPT Image 2 APIs, including batches and streamed previews.
When to use it
Use when a one-shot image helper is insufficient for image generation, masked edits, batches, streaming, or mixed text-and-image workflows.
Give it an image generation or editing request; it saves the resulting images to the chosen output paths.
Requires Python 3.10 or newer; the script uses only the standard library.
Requires an endpoint serving gpt-image-2; Responses workflows also require a text-capable Responses model such as gpt-5.4.
This skill
OpenAI
Sends prompts and images to OpenAI (irreversible)
.env
Reads configuration from parent directories