Image Generation

The Image Generation feature lets users create PNG images from text prompts, rendered directly in the chat and available for easy download. ASSIST AI currently supports five image generation models via OpenAI, Azure, or Vertex (Google). Support for additional providers such as Stability AI is planned for the future.

Model	Description
GPT-Image-1.5	Newest model — best for production-quality output with strong prompt control and reliable text/structured layouts
GPT-Image-1	Newer model with better prompt adherence — recommended as the default for most use cases
Nano Banana	Fast Gemini-based model optimized for quick, creative exploration and simple tasks
Nano Banana Pro	Advanced Gemini model delivering high-fidelity 4K output — ideal for complex tasks requiring consistency and detail
DALL-E 3	Older model — better suited for imaginative, richly stylized generations

Usage

Set up your preferred image generation models in the Admin Panel under:

Configuration → Image Generation

Once configured, Image Generation can be invoked from any Agent or AI model within ASSIST AI.

Getting Started

Core Features

Usage

​Usage

Usage