MiniMax Releases Video Model, Tops Both Global and Domestic Growth Charts on AI Product Ranking's September List | Oasis Vitality

Five weeks ago, MiniMax released its video model and launched it on Hailuo AI, unlocking a flood of limitless creativity. Over the past month, Hailuo AI's web traffic has surged more than 800%, landing it at #1 on both the global and domestic growth charts for AI products (web) in September. AI creators, filmmakers, and screenwriters from over 180 countries and regions worldwide are using Hailuo AI's magic to constantly push and expand the boundaries of creation. Oasis Capital has been alongside MiniMax since its Series A, firmly believing in the company's technical strength and creative power. Below is a showcase of MiniMax's video model achievements. Enjoy.

Currently, MiniMax's video model ranks #1 overall in third-party independent test results on VBench (a video generation model evaluation framework), leading across multiple dimensions including image quality, coherence, and fluidity. Hailuo Video has also added an "image-to-video" creation feature, and we look forward to more young creators joining in with new ways to play.

Using images as our creative foundation, we turn Hailuo's magic gear and open the door to parallel artistic universes. Everything, everywhere, all at once.

You can try out the new features on the Hailuo AI web version — unleash your wildest ideas and wander freely through the world of art. No waiting, experience it now:

For creators in China: https://hailuoai.com/video

For creators overseas: https://hailuoai.video/

New "Image-to-Video" Entry Point — Creation Experience Leveled Up

In Hailuo Video, users can click the "Image" button to freely upload an image as the first frame of their generated video, opening the door to a world of magic.

The MiniMax video model can accurately recognize uploaded images and ensure that the generated video maintains high consistency with the original input image in terms of appearance, while seamlessly embedding lighting and color tones into new scene settings — providing creators with space for coherent, in-depth creation.

Prompt (Image-to-Video): A man lifts his leg, steps over an advertising screen, and walks out of the building.

Prompt (Image-to-Video): A puppy jumps out of the picture frame and onto the desktop.

In terms of prompt following, the MiniMax video model can understand text that goes beyond the image content itself, deconstructing instruction frameworks and deep semantics to integrate them into video generation — capturing every subtle creative intention and achieving "what you write is what you see." With Hailuo AI's powerful image information control, every creator can be the most willful director.

Prompt (Image-to-Video): Right as the two people are arguing, the man suddenly transforms into a giant duck.

Prompt (Image-to-Video): A kitten breathes fire from its mouth, and the fish in front of it gets charred.

Special effects are no longer Hollywood's exclusive domain. Hailuo Video doesn't use effect templates — it relies purely on the model's comprehensive capabilities to achieve top-tier cinematic effects. Every user can fully unleash their imagination based on an image, creating richly varied movie-grade videos — CG compositing, scene transformation, fragmentation, anthropomorphism, and more effects and gameplay await your exploration.

Transformation Effect (Image-to-Video): The camera follows a large truck speeding down the highway, with dynamically blurred background. Suddenly, the truck transforms into a robot mid-drive.

CG Compositing (Image-to-Video): Transformers enter the frame and embrace two elderly people.

Object Anthropomorphism (Image-to-Video): Magic descends upon a plaster angel, bringing it to life.

Animal Anthropomorphism (Image-to-Video): A teddy bear walks into the frame, pats a smaller bear holding a sign, and they share a friendly hug.

Scene Transformation (Image-to-Video): Transform the city into a sci-fi scene style.

Shattering Effect (Image-to-Video): A man turns to look at the camera, his body splitting into countless fragments that fall to the ground.

Everyone Is an Artist — The Hailuo Vision Has Boundless Potential

Imagine yourself sitting in the director's chair, watching the monitor, walkie-talkie in hand, facing a cinematic dream factory woven from code. This small video model holds tremendous magic, letting everyone become an artist.

The passion and creativity everyone has shown for AI video creation has far exceeded our imagination. We've selected some of the most discussed and beloved video clips from social platforms at home and abroad — X, Discord, Weibo, WeChat, and more — hoping to spark some creative inspiration.

(The following videos are all user-created text-to-video works)

For musician Zachary Nelson, Hailuo AI-generated videos are the perfect companion to his AI-composed music. Through Hailuo AI, Zachary Nelson created an MV for his previously composed AI track "I Feel Your Pain," re-arranging and editing it for release on YouTube and Spotify.

How to go from laughing out loud to covering your face in tears within five seconds? Emotional expression has long been a challenge for AI, but Hailuo AI is like an acting coach. Singaporean creator el.cine used Hailuo AI to generate a close-up of a woman's face (T2V) with expression control rivaling professional actors. The single clip garnered nearly 300,000 views, with hundreds of commenters copying the prompt to join the "expression management" masterclass.

el.cine is also a film producer. He experimented with stitching together multiple Hailuo AI text-to-video creations, pairing them with music to generate a fast-paced short film with tense atmosphere. In his view, "AI is truly revolutionising the way we create."

Skateboarding enthusiast Lyre_e has 36 years of skating experience. Through Hailuo AI text-to-video, he created a series of 90s East Coast-style skate clips, reworking them into Street sk(AI)teboarding. With accurate recreation of 90s fashion and architecture, plus precise control of skateboarding movements through a distinctive fisheye lens, Hailuo AI earned praise on Reddit. Creator Lyre_e said, "They are good memories that some of us lived in skateboarding."

For newcomers who had never touched AI before, Hailuo AI is friendly and easy to pick up. Artist Ksenia_Galushkina used AI to create video for the first time — in fact, it was her first time independently creating video, period. Ksenia spent 2.5 days generating over 80 videos through Hailuo AI, combining them with post-production voiceover to complete the art film Dead & Breakfast Hotel, which she shared on Reddit as a Halloween creative experiment.

In the imagination of YouTube creators, wild animals are also good neighbors living in towns. A gorilla looks at the scenery from inside a car, a tiger wanders through the pantry, and monkeys cool off in the refrigerator. In the living room, there's even a group of baby rhinos running around and a baby elephant sneaking popcorn. AI enthusiasts commented, "The cutest AI video I've seen!" "This would be a wonderful utopia to live in."

User "anothermartz" used AI video on Reddit to parody and "expose" Hailuo AI's behind-the-scenes working principle — after receiving a user's video production request, the MiniMax team would have backstage staff build sets, wear masks, shoot, and edit to fulfill the request, then upload the finished product to the platform. This entire "play within a play" style piece, created entirely through text-to-video, was widely beloved. The realism of the staff "putting on a mask" to become celebrity chef Gordon Ramsay stunned all viewers.

In Hailuo Video's WeChat community, we hand the key to spacetime to every creator. In the first MiniMax Hailuo Video Prompt Challenge with the theme of "Time Travel," Hailuo AI opened the door to traversal, letting different eras intersect and collide. Here, cats transform into monsters crossing boundaries, waves crash upon puppies sleeping in living rooms; a youth practicing horsemanship dons armor to become a warrior, and a robot and a coyote share a cross-temporal handshake and embrace...

First "Prompt Challenge" Demo Video

Creator Submission Highlights

How to Play with Hailuo Video in Creative Ways — More Possibilities Await

MiniMax continues to invest in large video model R&D. At the same time, we look forward to co-creating intelligence with every user, jointly refining a better Hailuo. With each video generation, please give Hailuo a bit more patience and care, understanding and respecting AI's capabilities and limitations, so we can create more exquisite works together.

Recently, Hailuo AI Video also launched a prompt optimization feature. When enabled, the model automatically optimizes user prompts to improve video generation quality. When you don't have specific requirements for composition, atmosphere, movement, or camera work, we recommend turning this on — clever Hailuo AI will expand upon your original prompt to enhance the video's aesthetic presentation.

We've brought detailed prompt-writing tips for beginners — click "Read Original" to learn how to generate and edit high-quality AI shorts, making every frame an artistic showcase. Outstanding works will have the opportunity to be featured on the Hailuo Video homepage.

"Image-to-video" is our new step forward. Going forward, MiniMax will bring more creative video possibilities. Thank you to everyone who has provided feedback and suggestions on the video model, and we welcome more AI creators to share your exclusive creative secrets in the comments — let's exchange ideas and improve together.

More excitement, all on Hailuo AI.