GPT UI Icon Generator

A Unity Editor tool to generate clean UI icons using ChatGPT + DALL-E

What It Does

This tool refines icon prompts using ChatGPT and generates beautiful, white-background UI icons using OpenAI's DALL-E. It also removes backgrounds automatically with rembg, saving you clean Unity Sprites in one click.

Key Features

  • Refines icon prompts using ChatGPT for better results
  • Generates beautiful, white-background UI icons using OpenAI's DALL-E
  • Automatically removes backgrounds with rembg technology
  • Saves clean Unity Sprites in one click
  • Seamless integration with Unity Editor
  • Batch processing capabilities
  • Customizable prompt templates
Try on Unity Asset Store

How It Works

1

Refine with ChatGPT

Enter your icon concept and let ChatGPT optimize the prompt for better results.

2

Generate via DALL-E

The refined prompt is sent to DALL-E to create a high-quality icon image.

3

Save as Sprite

Background is automatically removed and the clean icon is saved as a Unity Sprite.

AI 3D Model Generator

Unity integration for AI-powered 3D model generation from text prompts and images

What It Does

A comprehensive Unity Editor tool that integrates with Tripo3D's AI-powered 3D generation API. Create 3D models from text descriptions or images and import them directly into your Unity project with full editor integration and runtime support.

Key Features

  • Text-to-3D model generation with customizable prompts
  • Image-to-3D conversion (single image or 4-view multiview)
  • Direct Unity Editor integration with GLB import
  • Runtime model spawning with glTFast support
  • Real-time progress tracking with Server-Sent Events
  • Configurable mesh complexity and texture resolution
  • Automatic asset management and prefab creation

Requirements

  • Unity 2020.2 or newer
  • Valid Tripo3D API key
  • glTFast package (for runtime spawning)
Try on Unity Asset Store

Generation Modes

Text → Model

Describe your 3D object in natural language

Image → Model

Convert single images to 3D models

Multiview → Model

Use 4 images (front, left, back, right) for precise reconstruction