MiniMax launches the video-01 video generator in Hailuo AI

Hailuo AI is MiniMax's consumer-facing AI platform. When it launched in March 2024 it handled text and music generation ^[1]. Six months later the platform gained a video-generation model, video-01, which became the product's best-known feature and, later, the subject of a copyright lawsuit in the United States.

Background

Hailuo AI started as a tool for generating text and music ^[1]. Video generation is a different problem: instead of predicting the next word or the next note, a model has to produce a sequence of images that stay consistent from frame to frame, so that motion looks continuous rather than like a slideshow of unrelated pictures. Building that capability into an existing platform meant adding a second kind of model alongside the text and music functions.

What happened

In September 2024 MiniMax released video-01 under the Hailuo AI platform ^[1]. KrASIA reported on 11 September 2024 that the company had launched the new model in early September, without giving an exact day ^[2], so public reporting does not pin the release to a single date.

Key details

According to KrASIA, video-01 could generate six-second clips at 1280x720 resolution and 25 frames per second ^[2]. Those numbers describe the output precisely. 1280x720 is a widescreen format, roughly the resolution families of a few years ago called 720p. Twenty-five frames per second means the model produces 25 still images for every second of video, which is the rate at which the sequence is played back to look like motion. A six-second clip at that rate is 150 generated frames that have to hang together.

Six seconds is short: long enough for a single shot or a brief action, not for a scene with cuts. The resolution and frame rate together set the practical use — short clips made to be viewed on a screen rather than material for broadcast or cinema.

Afterwards

Hailuo AI kept changing after the video model arrived. In January 2025 audio functions were added to the platform ^[1]. In December 2024, the trade publication Broadcast reported that Hailuo AI could reproduce British broadcasters' logos ^[1], a report that put the model's outputs in front of an industry audience that pays close attention to branding and rights.

In September 2025, Disney, Universal and Warner Bros. Discovery filed a copyright suit in the United States over Hailuo AI ^[1]. The suit turned a product question — what a video model can be asked to make — into a legal one about what such a system may produce and from what it learned.

Why it matters

video-01 marks the point where Hailuo AI grew from a text and music tool into a platform that also makes moving images ^[1]. It also illustrates a pattern that became common for generative video: a model is launched quickly, its limits are stated in technical terms such as clip length and frame rate ^[2], and the argument over rights and permitted outputs arrives afterwards, in this case from three major Hollywood studios ^[1].