Minimax Prompt Examples

Minimax, a new AI video generator has been released by the Chinese startup Hailuo AI.

At the time of writing the tool is currently free to use with no limits on video generation.

No doubt this is not going to be for long so if you're reading this, consider yourself lucky and seize your chance to try it out https://hailuoai.video ! You only need your email to sign up now.

👉 CHECK OUT MY HAILUO AI TUTORIAL

The model is only offering a text to video option at the moment with 500 2000 characters limit. It pops out 6 second long clips. No extensions or image to video. The generation typically takes a couple of minutes and you can not queue any additional video prompts until your current one finishes generating.

So far the video generation results suggest Minimax has very good prompt adherence, understands many styles, renders motion very well, and is pretty good with text. It also can generate celebrity videos. 

 I've translated the message on my queued video, it read "There are 39 people ahead, and the waiting time is expected to be 3 minutes".

Minimax Prompt Examples

Hailuo AI x Minimax understands various prompt formats, you don't need to worry about gettign it right. You can be vague or specific, and you can use various styles and formulas of prompting.  Here are just some of the prompts examples for Minimax:

Color Splash Prompt Example

In a monochrome black and white city street, full of monochrome people and cars in motion blur, focus on the single little girl in color. The girl is wearing a yellow dress and red shoes, she is standing in the middle looking up as her purple balloon is slowly rising up into the grey sky.

We can see that prompt was understood relatively well (girl keeps walking not standing, and holding the balloon, it's not flying off into the sky), the color on the girl looks painted on top in haste (monochrome edges on her dress), but check out the movements: our subject is actually walking at quite a pace! No slow-motion stuff as usual with current video generators. I like where this is going.

Youtuber's Livestream Video

A Youtuber doing a live stream in his home office, talking and gesticulating, smiling. Behind him there's a Youtube Creator Awards for Subscriber Milestones.

Here you can see how impressively realistic the generated video is. The movements, fingers are correct numbers and true to anatomy. It has a clear understanding of the concept of 'youtuber', 'influencer', how the Youtube awards look. This model rivals Sora and Runway.

Yellow Car Wheels Spin Prompt

Medium shot of the yellow car, tracking shot following its movement on the reflective surface, highlighting the reflections and neon lights.

Girl in a Milk Sea Text to Video Prompt

Surreal fantasy world: light pink sky, ocean made entirely of milk, focus on a 20 year old woman emerging from it after diving, as she looks at me. Woman's skin is smooth, white and moist, she has shoulder-long platinum blonde wet hair, the woman has slightly otherworldly appearance. Eyes deep green color almost glowing. White liquid surrounding her because surreal ocean is composed of white milk.

Cat Exorcism Text to Video Prompt

From below shot of a cat catholic priest performing an exorcism on a demonic cat, a parody on The Exorcist movie.

The expected result was actually a cat-priest, like on the image below which was generated by FLUX from the same exact prompt, but Minmax still done a great job with the overal atmosphere and characters.

BTW the word 'exorcism' is banned in Kling altogether, lol. Can't generate same prompt there, but able to upload flux's image and animate it without explicitly stating the word.

Spaceship Landing Text to Video

Wide shot of the alien planet, tracking the spaceship landing, capturing the futuristic and otherworldly environment.

Text Prompt + Pixar Style

Zoom in on a cute red shiny robot holding a white banner with text 'AGEofLLMs.com' standing in the big data center, in the Pixar animation style.

Tracking Man in Coat Prompt

From below tracking shot of a tall man from his back walking down a long symmetrical alley with trees on both sides. The man is wearing a long bright blue colored trench coat. The trees are magnificent boasting wide branches filled with abundant yellow leaves. It is windy so the man's coat and leaves on the trees are billowing, moving.

The model ignored the 'view form the back' aspect, this was actually the intended result for my prompt:

Last modified 24 September 2024 at 19:06