OpenAIImageGenerator

Generate images from text prompts using OpenAI image generation models. The component accepts a text prompt and returns generated images along with any revised prompt OpenAI used.

Key Features

• Uses gpt-image-2 by default.
• Supports configurable image quality and size.
• Returns images as base64-encoded JSON strings.
• Returns the revised prompt OpenAI used for image generation.
• Compatible with PromptBuilder for dynamic prompt construction.

Configuration

Authentication

To use this component, connect Haystack Platform with OpenAI first. For detailed instructions, see Use OpenAI Models.

1. Drag the OpenAIImageGenerator component onto the canvas from the Component Library.
2. Click on the component to open the configuration panel.
3. On the General tab:
   1. Select the image generation model. The default is gpt-image-2.
   2. Choose the image quality: auto, high, medium, or low.
   3. Set the image size. The standard options are 1024x1024, 1024x1536, 1536x1024, or auto.
4. Go to the Advanced tab to configure the OpenAI API key, timeout, maximum retries, API base URL, organization ID, and HTTP client settings.

Connections

OpenAIImageGenerator accepts a prompt string as input. It outputs a list of images (images) and the revised prompt (revised_prompt) OpenAI used for generation.

Typically, you connect a PromptBuilder to the OpenAIImageGenerator's prompt input to build dynamic prompts. The images and revised_prompt outputs connect to an OutputAdapter or AnswerBuilder for further processing.

Source Code

To check this component's source code, open openai_image_generator.py in the Haystack repository.

Usage Examples

Basic Configuration

  openai_image_generator:
    type: haystack.components.generators.openai_image_generator.OpenAIImageGenerator
    init_parameters:
      model: gpt-image-2
      quality: auto
      size: 1024x1024
      timeout: 60

Image Generation Pipeline

This pipeline uses OpenAIImageGenerator to generate images based on a query:

# haystack-pipeline
components:
  prompt_builder:
    type: haystack.components.builders.prompt_builder.PromptBuilder
    init_parameters:
      template: '{{query}}'
  openai_image_generator:
    type: haystack.components.generators.openai_image_generator.OpenAIImageGenerator
    init_parameters:
      model: gpt-image-2
      quality: auto
      size: 1024x1024
      timeout: 60
  answer_formatter:
    type: haystack.components.converters.output_adapter.OutputAdapter
    init_parameters:
      template: |-
        {% set ns = namespace(doc_string='') %}
        {% set ns.doc_string = ns.doc_string + '## Query:\n' + query + '\n\n' %}
        {% set ns.doc_string = ns.doc_string + '## OpenAIs Revised Prompt:\n' + revised_prompt + '\n\n' %}
        {% set ns.doc_string = ns.doc_string + '![](' + images[0] + ')' + '\n\n' %}
        {% set answer = [ns.doc_string] %}
        {{ answer }}
      output_type: List[str]
  answer_builder:
    type: deepset_cloud_custom_nodes.augmenters.deepset_answer_builder.DeepsetAnswerBuilder
    init_parameters: {}

connections:
- sender: prompt_builder.prompt
  receiver: openai_image_generator.prompt
- sender: openai_image_generator.revised_prompt
  receiver: answer_formatter.revised_prompt
- sender: openai_image_generator.images
  receiver: answer_formatter.images
- sender: answer_formatter.output
  receiver: answer_builder.replies
- sender: prompt_builder.prompt
  receiver: answer_builder.prompt

max_runs_per_component: 100

metadata: {}

inputs:
  query:
  - prompt_builder.query
  - answer_formatter.query
  - answer_builder.query

outputs:
  answers: answer_builder.answers

Parameters

Inputs

Parameter	Type	Default	Description
prompt	str		The prompt to generate the image.
size	Optional[Literal['1024x1024', '1024x1536', '1536x1024', 'auto']]	None	If provided, overrides the size provided during initialization.
quality	Optional[Literal['auto', 'high', 'medium', 'low']]	None	If provided, overrides the quality provided during initialization.
response_format	Optional[Literal['b64_json']]	None	Kept for backward compatibility. The component always returns a base64-encoded image.

Outputs

Parameter	Type	Description
images	List[str]	A list of generated images as base64-encoded JSON strings.
revised_prompt	str	The prompt OpenAI used to generate the image. OpenAI may revise the original prompt before generating the image.

Init Parameters

These are the parameters you can configure in Pipeline Builder:

Parameter	Type	Default	Description
model	str	gpt-image-2	The OpenAI image generation model.
quality	Literal['auto', 'high', 'medium', 'low']	auto	The quality of the generated image.
size	Literal['1024x1024', '1024x1536', '1536x1024', 'auto']	1024x1024	The size of the generated image. gpt-image-2 also supports custom sizes.
response_format	Literal['b64_json']	b64_json	Kept for backward compatibility. The component always returns a base64-encoded image.
api_key	Secret	Secret.from_env_var('OPENAI_API_KEY')	The OpenAI API key to connect to OpenAI.
api_base_url	Optional[str]	None	An optional base URL.
organization	Optional[str]	None	The Organization ID, defaults to None.
timeout	Optional[float]	None	Timeout for OpenAI Client calls. If not set, it is inferred from the OPENAI_TIMEOUT environment variable or set to 30.
max_retries	Optional[int]	None	Maximum retries to establish contact with OpenAI if it returns an internal error. If not set, it is inferred from the OPENAI_MAX_RETRIES environment variable or set to 5.
http_client_kwargs	Optional[Dict[str, Any]]	None	A dictionary of keyword arguments to configure a custom httpx.Clientor httpx.AsyncClient. For more information, see the HTTPX documentation.

Run Method Parameters

These are the parameters you can configure for the component's run() method. This means you can pass these parameters at query time through the API, in Playground, or when running a job. For details, see Modify Pipeline Parameters at Query Time.

Parameter	Type	Default	Description
prompt	str		The prompt to generate the image.
size	Optional[Literal['1024x1024', '1024x1536', '1536x1024', 'auto']]	None	If provided, overrides the size provided during initialization.
quality	Optional[Literal['auto', 'high', 'medium', 'low']]	None	If provided, overrides the quality provided during initialization.
response_format	Optional[Literal['b64_json']]	None	Kept for backward compatibility. The component always returns a base64-encoded image.

Related Information

• Use OpenAI Models
• Pipeline Components