Implementation:Openai Openai python Images Generate

Knowledge Sources	openai-python OpenAI Image Generation API
Domains	Computer_Vision, Image_Generation
Last Updated	2026-02-15 00:00 GMT

Overview

Concrete tool for generating images from text prompts with streaming support provided by the OpenAI Python SDK.

Description

The Images resource provides a generate() method that creates images from text descriptions. It supports DALL-E 2, DALL-E 3, and GPT-Image-1 models with configurable sizes, quality, formats, and streaming partial image previews.

Usage

Call client.images.generate() with a prompt and model. Access results via response.data[0].url or response.data[0].b64_json.

Code Reference

Source Location

• Repository: openai-python
• File: src/openai/resources/images.py
• Lines: L858-914 (sync impl), L1748-1804 (async impl)

Signature

class Images(SyncAPIResource):
    def generate(
        self,
        *,
        prompt: str,
        model: Union[str, ImageModel, None] | NotGiven = NOT_GIVEN,
        n: Optional[int] | NotGiven = NOT_GIVEN,
        quality: Literal["standard", "hd", "low", "medium", "high", "auto"] | NotGiven = NOT_GIVEN,
        response_format: Optional[Literal["url", "b64_json"]] | NotGiven = NOT_GIVEN,
        size: Optional[Literal["auto", "1024x1024", "1536x1024", "1024x1536", "256x256", "512x512", "1792x1024", "1024x1792"]] | NotGiven = NOT_GIVEN,
        style: Optional[Literal["vivid", "natural"]] | NotGiven = NOT_GIVEN,
        background: Optional[Literal["transparent", "opaque", "auto"]] | NotGiven = NOT_GIVEN,
        output_format: Optional[Literal["png", "jpeg", "webp"]] | NotGiven = NOT_GIVEN,
        output_compression: Optional[int] | NotGiven = NOT_GIVEN,
        partial_images: Optional[int] | NotGiven = NOT_GIVEN,
        stream: Optional[Literal[False]] | Literal[True] | NotGiven = NOT_GIVEN,
        moderation: Optional[Literal["low", "auto"]] | NotGiven = NOT_GIVEN,
        user: str | NotGiven = NOT_GIVEN,
    ) -> ImagesResponse | Stream[ImageGenStreamEvent]:
        """
        Creates an image given a prompt.
        """

Import

from openai import OpenAI
# Access via client.images.generate()

I/O Contract

Inputs

Name	Type	Required	Description
prompt	str	Yes	Image description
model	str	No	dall-e-2, dall-e-3, or gpt-image-1
n	int	No	Number of images (1-10)
size	str	No	Dimensions (e.g., "1024x1024")
quality	str	No	Quality level
response_format	str	No	"url" or "b64_json"
stream	bool	No	Enable streaming partial images
background	str	No	transparent, opaque, auto

Outputs

Name	Type	Description
response	ImagesResponse	Response with .data list of Image objects
stream	Stream[ImageGenStreamEvent]	Streaming partial image events

Usage Examples

Basic Generation

from openai import OpenAI

client = OpenAI()
response = client.images.generate(
    model="dall-e-3",
    prompt="A white cat sitting on a windowsill at sunset",
    size="1024x1024",
    quality="hd",
    n=1,
)
image_url = response.data[0].url
print(image_url)

Base64 Response

import base64

response = client.images.generate(
    model="gpt-image-1",
    prompt="A futuristic city skyline",
    response_format="b64_json",
    output_format="png",
)

image_bytes = base64.b64decode(response.data[0].b64_json)
with open("city.png", "wb") as f:
    f.write(image_bytes)

Streaming with Partial Previews

stream = client.images.generate(
    model="gpt-image-1",
    prompt="A detailed landscape painting",
    stream=True,
    partial_images=3,
)

for event in stream:
    if hasattr(event, "data") and event.data:
        print(f"Partial image received ({len(event.data[0].b64_json)} chars)")

Related Pages

Implements Principle

• Principle:Openai_Openai_python_Image_Generation

Requires Environment

• Environment:Openai_Openai_python_Python_3_9_Plus