The Best AI Tool For Coloring Pages | Midjourney vs Dall E 3 vs Leonardo AI

With so many AI tool out there and new ones popping up all the time.
I wanted to find the best one for making coloring pages.
I spent hours testing out the contenders to see which ones really stand out.
I focused on three main tools: Midjourney, Dall-E 3, and Leonardo AI.
I have subscriptions to all three and decided to put them to the test over five exciting rounds, each with a different theme, to see which AI tool could create the best coloring pages.
ROUND 1 — ANIMALS
Prompt: coloring page of a playful kitten chasing a ball of yarn in the kitchen
MIDJOURNEY

Midjourney gave me consistent images matching the prompt including the cat, yarn and kitchen background but missed capturing the kitten’s playful spirit with the yarn. I didn’t feel the images were particularly interesting.
DALL-E 3

Dall-E 3, produces one page at a time so to keep in consistent with other tool, I regenerated the same prompt 3 more times to produce 4 images.
I believe it nailed the prompt with playful and detailed images, despite a minor hiccup with the misconfigured chair and yarn attached to the cat in the fourth image.
LEONARDO AI

After hours of testing, I’ve found these to be the best settings to generate coloring pages in Leonardo.

Leonardo AI’s attempt had some grayscale issues, I struggled with this tool to produce consistent images the most. It sometimes produces colored images despite the negative prompts.
It did generate the cat, yarn and kitchen but the images don’t look particularly playful or show the cat with any emotion.
The Winner
Dall-E 3 is the winner of this round with its fun and accurate portrayal as it was closet to matching the prompt.
ROUND 2 — A PERSON
Prompt: coloring page of a firefighter in uniform, standing next to a fire engine
MIDJOURNEY

Midjourney impressed me with detailed images and a creative touch, even including a firefighter holding a hose.
It stuck closet to the prompt with realistic images.
DALL-E 3

Dall-E 3 really struggled with this one, with too much grayscale and missing faces.
LEONARDO AI

Leonardo AI also had a tough time, especially with the grayscale. It didn’t look anything like a fire engine behind the firefighter in the first image.
I would only use the fourth image in a coloring book as the rest were far below quality.
The Winner
Midjourney was the clear winner here, showing off its detailed and creative for people.
ROUND 3 — LANDSCAPE
Prompt: coloring page of a tropical beach with palm trees leaning over the shore and few small boats in the distance
MIDJOURNEY

Midjourney delivered consistent quality but produced a small boat on the beach in 2 of the images instead of them all being in the distance as written in the prompt.
DALL-E 3

Dall-E 3 stuck to the prompt well, creating scenes with trees over the shore and boats in the distance.
LEONARDO AI

Leonardo AI improved but still faced challenges with grayscale.
Only in the fourth image, I find the boats to be in the distance.
The Winner
Dall-E 3 won this round for sticking closest to the prompt and impressing me with its detailed pages.
ROUND 4 — QUOTES
Prompt: a coloring page with the words ‘Believe in Yourself’ in white with a black outline surrounded with small, simple flowers and leaves that can be colored in
MIDJOURNEY

Midjourney had trouble with the words and the black outline.
It isn’t quite there yet with the text capabilities but it will definitely improve in the coming month.
DALL-E 3

Dall-E 3 set itself apart despite some small spelling errors, it beautifully blended the words with a floral background.
LEONARDO AI

Leonardo AI did a better job than Midjourney, it’s not known for it’s text capabilities but it wasn’t a bad attempt.
The Winner
The quote coloring pages generated by Dall-E 3 were far above the rest making it the winner of this round.
ROUND 5 — MANDALA
Prompt: a coloring page of a floral-inspired mandala, featuring layers of petals, leaves, and blossoms radiating from the center
MIDJOURNEY

Midjourney produced flawless mandalas that perfectly matched the prompt.
I love how the detail isn’t lost even when you take a closer look at the Mandala.

DALL-E 3

While Dall-E 3’s images look good from afar.
The detail is lost up close and some areas become blurry.

LEONARDO AI

Leonardo AI’s attempt was far off the mark. Too much greyscale and some of the lines are all over the place.
The Winner
Midjourney took this round, impressing me with the quality and detail of its mandalas.
Final Thoughts
Overall, Dall-E 3 emerged as my favorite, winning three out of five rounds with its ability to closely follow prompts and create playful, detailed images.
Midjourney cane close second producing the most realistic images and Leonardo finished third as it didn’t have the quality to compete with the others.
I’d love to hear your thoughts or if you’re interested in a similar comparison for book covers.

This story is published on Generative AI. Connect with us on LinkedIn and follow Zeniteq to stay in the loop with the latest AI stories. Let’s shape the future of AI together!






