MiniMax releases Hailuo 2.3 video model
MiniMax kept pricing level with the prior Hailuo 02 model while improving character movement, facial micro-expressions and stylised rendering.
- Models & capabilities
- Colour
MiniMax released Hailuo 2.3, an update to its text- and image-to-video generation model, improving on its predecessor in physical movement, facial micro-expressions and stylised rendering — including anime, ink-wash painting and game cinematics. The company said it held pricing at prior Hailuo 02 levels, framing the release as more capability for the same cost rather than a price increase, alongside a faster, cheaper “Fast” variant aimed at cutting batch-generation costs.
The release also folded MiniMax’s Hailuo Video Agent into a broader “Media Agent,” which selects an appropriate underlying model automatically from a user’s prompt rather than requiring manual model selection, with an option to customise generation step by step.
The update was one of several from Chinese video-generation labs through 2025 competing on cost and iteration speed against Western entrants such as Runway and Google’s Veo, rather than on a single headline capability jump.