July 31, 2026
Lights, camera… slop?
What Is MiniMax H3? Everything You Need to Know About the Hailuo 3.0 Video Model
MiniMax says its new AI makes movie-like clips, but commenters are already yelling “slop”
TLDR: MiniMax launched H3, a new AI video tool that can make short clips with sound in one go, aiming to turn simple prompts into mini scenes. But the early community mood is icy, with critics instantly dismissing it as “AI slop” and questioning whether better specs actually mean better content.
MiniMax has rolled out H3, also called Hailuo 3.0, and on paper it sounds like a flashy leap for AI video: short clips in sharp 2K quality, built-in sound, and up to 15 seconds of footage generated from a prompt, picture, or other reference material. The company is pitching it as more than a silent animation toy. It wants to be a one-stop machine for making tiny scenes complete with voices, background noise, and cinematic framing. In plain English: type what you want, and it tries to spit out a mini movie scene.
But the real action is in the reaction, and the mood so far is brutal, blunt, and very online. The loudest comment cuts straight to the chase: “Low quality ai slop”. That short jab basically hijacks the whole vibe around the launch. Instead of cheering the higher resolution and built-in audio, skeptics are rolling their eyes at yet another AI video tool promising the future while reminding people of weird, plastic-looking clips flooding social feeds. The drama here is simple and spicy: MiniMax is selling “film-like,” while critics are hearing “more algorithm soup.”
And yes, there’s a meme energy to it all. “Slop” has become the internet’s favorite insult for mass-produced AI content, so the comment lands like a roast, not just a review. That makes H3’s launch feel less like a clean victory lap and more like a public trial in the court of the comments. Fancy features are nice, but the community’s first question is much messier: is this actually useful, or just more polished nonsense?
Key Points
- •MiniMax launched MiniMax H3, also called Hailuo 3.0, on July 31, 2026 after previewing it at WAIC 2026.
- •H3 is a multimodal AI video model that generates native 2K, 24fps video with synchronized stereo audio from text, images, and other reference materials.
- •The model supports 5 to 15 seconds of generation per request, with extension to about 30 seconds using an Extend tool.
- •H3’s Omni-Reference feature allows up to 9 images, 3 video clips, and 3 audio clips to be used together to improve character, motion, and voice consistency.
- •The article lists pricing at about $0.13 per second at 2K and notes instruction-based editing for existing generations.