Package bg.codexio.ai.openai.api.sdk.images

Class AIModelStage<R extends bg.codexio.ai.openai.api.payload.images.request.ImageRequest,E extends RuntimeSelectionStage>

java.lang.Object
bg.codexio.ai.openai.api.sdk.images.ImageConfigurationStage<R>
bg.codexio.ai.openai.api.sdk.images.AIModelStage<R,E>
All Implemented Interfaces:
IntermediateStage
public class AIModelStage<R extends bg.codexio.ai.openai.api.payload.images.request.ImageRequest,E extends RuntimeSelectionStage> extends ImageConfigurationStage<R> implements IntermediateStage
A stage to choose a model supported by /v1/chat/completions model endpoint compatibility .

As for the current version of the API, this step is supported only for creating images. All other images related actions support dall-e-2 only.

  • Method Details

    • poweredByDallE2

      public ChoicesStage<R,E> poweredByDallE2()
      Configures the AI Model to be DallE20

      This model is the default model for images generation and the ancestor of the most recent dall-e-3 model. It can create original, realistic images based on other art or text description. Keep in mind that images may be more cartoon-ish looking and more similar to each other.

      Returns:
      ChoicesStage to configure the choices count.

    • poweredByDallE3

      public StyleStage<R,E> poweredByDallE3()
      Configures the AI Model to be DallE30

      This model is the most modern model for images generation and the successor of the default dall-e-2 model. It understands significantly more nuance and detail compared to the earlier systems. Creates accurate images by directly translating the provided instructions.

      Note: Currently supports only images generation and is slightly more expensive in tokens

      Returns:
      ChoicesStage to configure the choices count.

    • defaultModel

      public ChoicesStage<R,E> defaultModel()
      Configures the AI Model to be DallE20

      This model is the default model for images generation and the ancestor of the most recent dall-e-3 model. It can create original, realistic images based on other art or text description. Keep in mind that images may be more cartoon-ish looking and more similar to each other.

      Returns:
      ChoicesStage to configure the choices count.