Skip to main content
The Image as Output feature allows your custom Agent to generate images directly in response to a user’s prompt. This is especially powerful for Agents specializing in visual content, such as architectural visualization, graphic design, or creative art generation.

1. Enabling Image as Output

To enable this feature, navigate to the Core Features section while creating or editing your Agent and toggle the switch for Image as Output.
  • Go to Core Features: Locate the Image as Output toggle switch.
  • Enable the Feature: Switch the toggle ON.

2. Configuring the Image Provider

Once enabled, you must configure the specific image generation model your Agent will use. Click the Configure link to open the configuration panel.

Select Image Provider & Model

The platform allows you to choose from various integrated image generation services. Simply click the dropdown to see the available providers.

Google Image Models (Nano Banana Series)

If you select Google, you can choose from the specialized Nano Banana suite:
  • Nano Banana 2: Our latest state-of-the-art model for text-to-image accuracy and high-fidelity composition.
  • Nano Banana Pro: Optimized for professional-grade aesthetic quality.
  • Nano Banana: The standard, efficient model for rapid generation.

OpenAI Image Models

If you select OpenAI, you will have access to the latest DALL-E and GPT image models:
  • gpt-image-1.5: The newest iteration for highly contextual image generation.
  • gpt-image-1: Reliable, prompt-adherent image creation.
  • dall-e-3: High-detail, photorealistic generation.
  • dall-e-2: Efficient, versatile generation.
After selecting your preferred provider and model, click the Save button.

3. Agent Configuration Example

The screenshots illustrate a “Dream House Image Generator” Agent configured for image output:
FieldConfigurationPurpose
NameDream House Image GeneratorDescriptive Agent name.
ModelOpenAI / gpt-4tThe LLM driving the Agent’s reasoning.
Agent RoleExpert Architectural Visualization Assistant…Defines the Agent’s persona and expertise.
Agent GoalProduce accurate, high-quality images…Clear objective for the Agent’s output.
Agent InstructionsOverview: You generate house images on demand…Detailed instructions on output style, camera angles, and format.
Core FeatureImage as Output (Enabled)Activates the image generation capability.

4. Testing the Agent Inference

Once configured, the Agent processes text prompts and uses the selected provider to generate visual output.
User Prompt: “I want the front view of a modern french style palace. I want it to be in shaded of black and wooden color. It needs to be photo realistic one.”
The Agent responds by:
  1. Generating the image file (JPEG/PNG).
  2. Providing the refined prompt used by the underlying image model.
  3. Offering a direct download/view link.

Example Output

The resulting image generated by the Agent based on the user’s prompt: This feature empowers developers to build Agents capable of delivering complex, custom visual content directly within a conversational workflow.