# Art Generator
**URL:** https://community.glideapps.com/t/art-generator/66105
**Category:** Project Showcase
**Created:** [September 13, 2023, 7:39pm UTC](https://community.glideapps.com/t/art-generator/66105 "2023-09-13T19:39:42Z")
**Posts on this page:** 4
**Page:** 1
### Author:  [@IdeafideTeam](https://community.glideapps.com/u/IdeafideTeam)
#### Post date: [September 13, 2023, 7:39pm UTC](https://community.glideapps.com/t/art-generator/66105/1 "2023-09-13T19:39:42Z")
Here’s an awesome art generator I made that writes unique prompt options based on your input. It is a great tool for generating assets without needing to be an art expert or prompt engineer.
> **[Idea Art](https://ideafide-art.glide.page/dl/74c4fe)**
>
> An auto-prompter for DALLE
---
### Author:  [@ThinhDinh](https://community.glideapps.com/u/ThinhDinh)
#### Post date: [September 14, 2023, 1:57am UTC](https://community.glideapps.com/t/art-generator/66105/2 "2023-09-14T01:57:11Z")
Cool DALL-E use case. Is it true that you can only use styles that are pre-defined?
---
### Author:  [@IdeafideTeam](https://community.glideapps.com/u/IdeafideTeam)
#### Post date: [September 14, 2023, 8:12am UTC](https://community.glideapps.com/t/art-generator/66105/3 "2023-09-14T08:12:32Z")
**Everything is generated realtime - there are no predefined styles or prompts**
First, GPT-4 generates 4 prompt ideas based on the user’s idea. Each prompt is parsed into cards. Clicking on a card triggers an action that sends the prompt to DALLE2 to generate the image which is displayed below.
- It’s been a pretty efficient way to gen assets.
Let me know any other questions or suggestions. Happy to send over a template link.
thanks so much for trying it
---
### Author:  [@Onyxonic](https://community.glideapps.com/u/Onyxonic)
#### Post date: [April 10, 2026, 8:43pm UTC](https://community.glideapps.com/t/art-generator/66105/4 "2026-04-10T20:43:18Z")
Jumping in on this older post because I’m curious if you ever shared that template link. I’d love to try the setup myself. Also wondering if you’ve tested it with newer image models since then and whether the prompt-card flow still works smoothly.