AI Atlas AI ATLAS

Hailuo

A short-video model focused on instruction following and camera language

MiniMax Model Closed

inTextImageVideo

WHAT IT IS

Hailuo is a video-generation model MiniMax released in September 2024. It generates clips of about 6 seconds at 720p and 25 frames per second from text or an image, with a focus on prompt adherence and natural camera movement. It is offered to users through products such as Hailuo AI. The model does not release its weights.

Why it matters

It stood out in short-video generation for comparatively strong prompt adherence and fine camera motion, another example of a Chinese team joining the text-to-video race in 2024.

Key specs

Resolution: 720p
Duration: About 6 seconds
Frame rate: 25 fps
Input: Text, image
Open weights: No

Capabilities

Text-to-Video

Turn a sentence into a short video clip

Image-to-Video

Set a still image in motion

Related concepts

Diffusion Models

Learn a thousand tiny denoising steps, and you can build an image from pure noise

Multimodal Generation

One model that learns to speak, to draw, to move — even to model the 3D world

Comparable products

Sora

Generates coherent video up to about a minute long from a description

Kling

Kuaishou (Kling)

A short-video model for both text-to-video and image-to-video

Gen-3 Alpha

A highly controllable text-to-video model for film and advertising

Seedance

ByteDance (Seed)

A video-generation model aimed at multi-shot storytelling

Dream Machine

Generates short videos with motion from text or an image