Birds beat cats and dogs in Magic Hour AI video prompts
Quick Answer
Quick answer
Bird terms appeared in stored prompts from 2,921 Magic Hour text-to-video creators during June through August 2026, compared with 1,724 for dogs and 1,618 for cats. Butterfly terms also appeared for more creators than either pet.
The analysis covers 92,976 completed text-to-video projects from 48,971 creators. Bird terms reached 69% more creators than dog terms and 81% more than cat terms. These figures describe stored text associated with completed projects on Magic Hour, rather than the subjects verified in the finished videos or the popularity of videos with viewers.
A red ceramic mug sits on a wooden café table beside a window. Steam rises slowly while the camera makes a gentle push-in. Soft morning light, realistic materials, one continuous shot, no people or readable text.
Turn your own idea into a video
Write a short scene with the subject and action you want, generate one clip, and see how the prompt behaves.
Try Text to VideoWings ahead of paws
Bird: 2,921 creators, 5.96% of all creators, and 4,290 completed projects mentioning the terms.
Butterfly: 1,912 creators, 3.90% of all creators, and 2,924 completed projects mentioning the terms.
Dog: 1,724 creators, 3.52% of all creators, and 2,623 completed projects mentioning the terms.
Cat: 1,618 creators, 3.30% of all creators, and 2,618 completed projects mentioning the terms.
Creators can appear in several categories, so the rows are not mutually exclusive. This comparison covers four selected categories and is not an exhaustive ranking of all animals.
Butterfly terms appeared for 11% more creators than dog terms and 18% more than cat terms. That makes the butterfly comparison another useful detail for anyone tracking the imagery associated with AI video creation.
What the numbers can tell us
This sample suggests that familiar pet imagery is only one part of how Magic Hour's text-to-video tools are used. Birds and butterflies offer another subject for research into the visual culture forming around generative video.
Flight, movement and fantasy are possible explanations for the result, but this analysis does not establish why the terms appeared. It does not show whether an external social trend caused the pattern, whether users preferred one subject over another, or whether a finished video attracted views.
Methodology
The analysis covers qualifying completed Magic Hour text-to-video projects created from June 1 through August 31, 2026, using UTC dates. A creator is a distinct user associated with at least one qualifying project during that window.
The corpus excludes templates, automated free-tool jobs, failed generations, deleted or disabled projects, and blank prompts. Case-insensitive term matching identifies mentions in stored prompt text. Bird terms match bird or birds while excluding bird eye and bird-eye-view phrasing. Dog terms match dog, dogs, puppy, puppies, canine or canines. Cat terms match cat, cats, kitten, kittens, feline or felines. Butterfly terms match butterfly or butterflies.
Distinct creators are counted separately for each category. Relative comparisons divide the difference in creator counts by the comparison category count. For example, (2,921 - 1,724) / 1,724 produces the rounded 69% bird-versus-dog difference. Creator shares use the full denominator of 48,971.
Limitations and privacy
This is a sample of Magic Hour activity, not a representative survey of all AI video creators. The language and terms matched by the queries affect which projects are counted. Stored text may also reflect prompt enhancement or other product behavior, so it should not be treated as verified verbatim creator-authored text.
The analysis reports aggregate counts. It does not publish raw prompts, user identities, project identifiers, uploaded media or generated videos. Every displayed category contains at least 100 distinct creators.
Citation-ready finding
According to Magic Hour's analysis of 92,976 completed text-to-video projects from June through August 2026, bird terms appeared in stored prompts from 2,921 creators, 69% more than dog terms and 81% more than cat terms. Butterfly terms also exceeded both pets.
Source: Magic Hour, analysis of qualifying completed text-to-video projects, June 1 through August 31, 2026. Aggregate tables and research queries are available from the research team.
The same retained prompt dataset also found that dragons beat unicorns 7 to 1.
A separate slice found that teachers beat astronauts 4 to 1.
Runbo Li is the Co-founder and CEO of Magic Hour, where he builds AI video and image tools for content creation. He is a Y Combinator W24 founder and former Data Scientist at Meta, where he worked on 0-1 consumer social products in New Product Experimentation. He writes about AI video generation, AI image creation, creative workflows, and creator tools.
