MiniMax H3 Video Generation Model Joins LMSYS Video Arena
MiniMax H3, the latest flagship video generation model from MiniMax, has been added to the LMSYS Video Arena for text-to-video and image-to-video benchmarking. Additionally, the developers announced that the model's weights will be open-sourced soon. The inclusion of MiniMax H3 in the LMSYS Video Arena allows users to evaluate its performance against other top video models through crowdsourced blind tests. The promise of upcoming open weights is highly significant for the open-source AI community, which currently lacks many high-quality open video generation models. MiniMax H3 (also known as Hailuo 3.0) supports native 2K video generation at 24 frames per second, producing clips ranging from 5 to 15 seconds with native audio. The model functions as a multimodal system that processes text, images, video, and audio within a single context.
## BACKGROUND
LMSYS Org (Large Model Systems Organization) runs the Chatbot Arena and its video counterpart, which are popular benchmarking platforms where users compare outputs from two anonymous models side-by-side to vote on the better one. MiniMax is an AI research lab known for its Hailuo series of video generation models, which compete with other state-of-the-art video generators.