MiniMax H3

#3 in Text-to-Video

Minimax · 17× · tolest 05. Aug. 2026

88

Momentum

MiniMax H3 is a general-purpose omni-modal video generation model officially introduced by MiniMax on July 31, 2026. It processes text, images, videos, and audio together as context and generates videos with native stereo sound in up to 2K resolution and lengths of up to 15 seconds. The model is currently available via the MiniMax API and the Hailuo AI consumer platform; MiniMax announced plans to publicly release the model weights in the coming days under the MiniMax Community License. According to MiniMax and third-party reports, including the South China Morning Post, the price per second at 2K resolution is less than one-third that of comparable mainstream models.

Belege (17)