Image Generator

ImageGenerator generates Texture2D images from text prompts. It supports OpenAI (DALL·E 3, GPT-Image-1) and Google Imagen providers.

Component Menu

AI Dev Kit / AI Generators / Image Generator

Inspector Settings

Field	Description
Settings	ImageGenerationSettings asset — model, size, quality, style
Save Outputs	Persist generated images to the output folder
Output Path	Folder path for saved images

ImageGenerationSettings fields

Field	Description
Model	Image generation model (e.g. dall-e-3, gpt-image-1, imagen-3.0)
Size	Output image size (e.g. 1024x1024)
Quality	Standard or HD (OpenAI models)
Style	Vivid or Natural (DALL·E 3 only)
Aspect Ratio	Portrait / Landscape / Square (Google Imagen)
Person Generation	Allow or block human faces (Google Imagen)

Events

Event	Signature	Description
onOutput	UnityEvent<Texture2D>	Fires when the final image is ready.
onStream	UnityStreamEvent<Texture2D>	Fires on each streaming frame (if supported by model).
onStatusChanged	UnityEvent<ResponseStatus>	Fires on status change.

Methods

Method	Description
Generate(string prompt)	Fire-and-forget generation from a text prompt.
GenerateAsync(string prompt)	Awaitable version. Returns Generated<IImageAsset>.
Configure(ImageGenerationSettings)	Override inspector settings at runtime.

Code Example

using Glitch9.AIDevKit.Generators;
using Cysharp.Threading.Tasks;
using UnityEngine;
using UnityEngine.UI;

public class ImageGenExample : MonoBehaviour
{
    [SerializeField] private ImageGenerator imageGenerator;
    [SerializeField] private RawImage displayImage;

    private void Start()
    {
        imageGenerator.onOutput.AddListener(tex => displayImage.texture = tex);
    }

    public void OnGenerateButtonClicked(string prompt)
    {
        imageGenerator.Generate(prompt);
    }
}