Implementation:Openai Openai node Images Resource

Knowledge Sources	Openai_Openai_node
Domains	SDK, Images, Generation
Last Updated	2026-02-15 12:00 GMT

Overview

The Images resource class provides methods for generating, editing, and creating variations of images through the OpenAI API.

Description

The Images class extends APIResource and exposes three primary operations: generate for creating images from text prompts, edit for modifying existing images based on prompts and optional masks, and createVariation for generating variations of a given image. The resource supports multiple model families including GPT image models (gpt-image-1.5, gpt-image-1, gpt-image-1-mini) and DALL-E models (dall-e-2, dall-e-3).

Both generate and edit support streaming via method overloads. When streaming is enabled, the API returns a Stream of partial image events (ImageGenStreamEvent or ImageEditStreamEvent) that deliver progressive base64-encoded image data, culminating in a completed event with the final image and usage statistics. Non-streaming calls return an ImagesResponse containing an array of Image objects with optional base64 data, URLs, or revised prompts.

The module defines extensive TypeScript types for all parameters and response shapes, including configurable options for image size (1024x1024, 1536x1024, 1024x1536, etc.), quality levels, output formats (png, jpeg, webp), background transparency, and compression levels. File uploads are handled through the Uploadable type with multipart form encoding.

Usage

Use this resource when you need to generate images from text descriptions, edit existing images with inpainting or outpainting, or create variations of source images. The streaming mode is particularly useful for providing progressive rendering feedback in user interfaces.

Code Reference

Source Location

• Repository: openai-node
• File: src/resources/images.ts

Signature

export class Images extends APIResource {
  createVariation(body: ImageCreateVariationParams, options?: RequestOptions): APIPromise<ImagesResponse>;

  edit(body: ImageEditParamsNonStreaming, options?: RequestOptions): APIPromise<ImagesResponse>;
  edit(body: ImageEditParamsStreaming, options?: RequestOptions): APIPromise<Stream<ImageEditStreamEvent>>;

  generate(body: ImageGenerateParamsNonStreaming, options?: RequestOptions): APIPromise<ImagesResponse>;
  generate(body: ImageGenerateParamsStreaming, options?: RequestOptions): APIPromise<Stream<ImageGenStreamEvent>>;
}

Import

import OpenAI from 'openai';

I/O Contract

Inputs (generate)

Name	Type	Required	Description
prompt	string	Yes	Text description of the desired image (max 32000 chars for GPT image models)
model	ImageModel | null	No	Model to use: dall-e-2, dall-e-3, gpt-image-1, gpt-image-1-mini, gpt-image-1.5
n	null	No	Number of images to generate (1-10; dall-e-3 only supports 1)
size	'1024x1024' | '1536x1024' | '1024x1536' | ...	No	Image dimensions
quality	'hd' | 'low' | 'medium' | 'high' | 'auto'	No	Image quality level
output_format	'jpeg' | 'webp' | null	No	Output format (GPT image models only)
background	'opaque' | 'auto' | null	No	Background transparency (GPT image models only)
stream	null	No	Enable streaming mode for progressive rendering
partial_images	null	No	Number of partial images during streaming (0-3)
style	'natural' | null	No	Image style (dall-e-3 only)

Inputs (edit)

Name	Type	Required	Description
image	Array<Uploadable>	Yes	Source image(s) to edit (up to 16 for GPT image models)
prompt	string	Yes	Description of the desired edit
mask	Uploadable	No	Image mask indicating areas to edit (transparent areas)
input_fidelity	'low' | null	No	How closely to match input image features

Outputs

Name	Type	Description
created	number	Unix timestamp of when the image was created
data	Array<Image>	List of generated images with b64_json, url, or revised_prompt
usage	ImagesResponse.Usage	Token usage information (GPT image models)
background	'opaque'	Background parameter used
output_format	'webp' | 'jpeg'	Output format used
quality	'medium' | 'high'	Quality level used
size	'1024x1536' | '1536x1024'	Image dimensions used

Usage Examples

import OpenAI from 'openai';
import fs from 'fs';

const client = new OpenAI();

// Generate an image
const imagesResponse = await client.images.generate({
  prompt: 'A cute baby sea otter',
  model: 'gpt-image-1',
  size: '1024x1024',
  quality: 'high',
});

// Edit an image with a mask
const editResponse = await client.images.edit({
  image: fs.createReadStream('path/to/image.png'),
  prompt: 'A cute baby sea otter wearing a beret',
  mask: fs.createReadStream('path/to/mask.png'),
});

// Create a variation (dall-e-2 only)
const variationResponse = await client.images.createVariation({
  image: fs.createReadStream('otter.png'),
});

// Streaming image generation
const stream = await client.images.generate({
  prompt: 'A futuristic cityscape',
  model: 'gpt-image-1',
  stream: true,
  partial_images: 2,
});

for await (const event of stream) {
  if (event.type === 'image_generation.partial_image') {
    console.log('Partial image received:', event.partial_image_index);
  } else if (event.type === 'image_generation.completed') {
    console.log('Final image ready, tokens used:', event.usage.total_tokens);
  }
}

Key Types

Image

interface Image {
  b64_json?: string;
  revised_prompt?: string;
  url?: string;
}

ImageModel

type ImageModel = 'gpt-image-1.5' | 'dall-e-2' | 'dall-e-3'
  | 'gpt-image-1' | 'gpt-image-1-mini';

Stream Event Types

type ImageGenStreamEvent = ImageGenPartialImageEvent | ImageGenCompletedEvent;
type ImageEditStreamEvent = ImageEditPartialImageEvent | ImageEditCompletedEvent;

Related Pages

• Environment:Openai_Openai_node_Node_20_Runtime