Welcome back to my series of posts where I test AI image generators and see how they handle making maze art. I will be asking 10 prompts and seeing what gets generated. My goal is to evaluate different AI image sites against each other to see how they perform. Here is the series so far:
An exploration into Al Image Maze Generation
Making mazes with AI: DallE2 (Updated to compare 2023 to 2026)
Making mazes with AI: Stable Diffusion (Updated to compare 2023 and 2024 to 2026)
Making mazes with AI: Craiyon (Updated to compare 2023 to 2026)
Making mazes with AI: Dream by Wombo (Updated to compare 2023 to 2026)
Making mazes with AI: Nightcafe (Updated to compare 2023 to 2026)
Making mazes with AI: StarryAI (Updated to compare 2023 to 2026)
Making mazes with AI: DeepAI (Updated to compare 2023 to 2026)
Making mazes with AI: Canva (Updated to compare 2023 and 2024 to 2026)
Making mazes with AI: AI Image Generator (Updated to compare 2023 to 2026)
Making mazes with AI: Photosonic
Making mazes with AI: Bing AI Image Generator (Updated to compare 2023 and 2024 to 2026)
Making Maze Art with google gemini
2024: You can access the Gemini website here. You must sign into your google account to use Gemini. There is not limit to the number of images you can generate. Each prompt returns 2 images with the option to generate more. For this exercise I chose the one closest to what I had asked for, or the most interesting.
My original testing took place in March 2023. Gemini (formerly Bard) just started this feature in Feb 2024 so it late to the party.
2026: You still need to sign in to use the site. Just type and the prompt will use Non Banana to create an image for you.
Prompt 1 - Make a medium difficulty maze of the Eiffel Tower in black and white with arrows at the start and finish
2024: That is the Eiffel Tower. There is a maze looking section and an arrow. Typically I do not get arrows, so this is an improvement.
2026: When this first generated and I saw the arrows and the maze inside the Eiffel Tower I thought we had out first AI maze art generationā¦but not quite.
Prompt 2 - Draw a medium difficulty large maze of the Empire State Building with the start and goal embedded in the structure
2024: Similar to the 1st maze but in a new style. Got the famous building correct. Looks like a maze but is not a maze.
2026: So close again. The start and goal make no sense no I see the building and the maze.
Prompt 3 - Draw a difficult maze of the White House pixel art style
2024: Again, we have the right building. Wrong style. No Maze. But good job if I had written a different prompt.
2026: Similar to the previous new generations, the start and goal make no sense, but the building is correct and I see an internal maze !
Prompt 4 - Draw a difficult maze that looks like a drawing of a famous building in sketch style
2024: The top is a sketch of the very famous ______ building. The bottom is a hand drawn maze that actually is a maze, i mean if it had a start and goal !
2026: Great sketch of the Roman Coliseum in Rome with no maze, but a start and a goal.
Prompt 5 - Draw a maze in the style of doyoumaze.com of a skyscraper in NYC
2024: Cool NYC image. I think I like the marble sky. No Maze, not my style, but at least it is cool !
2026: Reminds me of the ESB above.
Prompt 6 - Draw a maze in the style of Sean C Jackson of a scene from a large outdoor market
2024: This has potential. Change the perspective and switch from watercolor to more comic and youāve got it.
2026: It is a fun comic style black and white market using circus tents. Has potential. Crowded.
Prompt 7 - Make a maze of a slice of an orange in color
2024: Sure.
2026: The start and finish make sense and the random arrow on the right side could be removed.
Prompt 8 - Make a maze integrated on top of a photograph of a king sitting on his throne looking cantankerous beside his beautiful queen
2024: Weird. I didnāt ask for a specific king and queen. Any would do. If I ask for a plumber or bus driver is it blocked ? And, after I asked thisā¦well you all saw the many stories !
2026: An overlay of an (almost) maze over a photoā¦I guess Google figured out that this prompt was fine.
Prompt 9 - Make a solvable maze that is very large and very difficult to solve because it is so complex
2024: Looks like a hedge maze from above. Nice style.
2026: A maze algorithm away from being great.
Prompt 10 - Make a 3d render of a red and blue glossy cube maze
2024: No.
2026: The old double start.
Summary and Conclusion:
2024: How did Google Gemini do ? I think it was average. It has less structure than other text to image generators. You can ask for whatever you want (except people I guess). And there are no limits. But it also has no built in styles or adjustment buttons that are helpful with other AI image sites.
2026: Although no generator has incorporated maze algorithms into their process I think Gemini fakes it best. Solid output across the board and one of the reasons I use Gemini most often when I need a background or to make a comic strip.
