MiniMax H3 Launches as Unified Multimodal AI Model, Generating Text, Images, Video, and Audio at 2K Resolution for a Third of Competitor Prices
Summary
MiniMax H3 launches as a groundbreaking unified multimodal AI model that generates text, images, video, and audio in a single system, producing 2K resolution videos at less than a third of competitor prices, with open-source model weights coming soon.
Key Points
- MiniMax H3 launches today as a general-purpose multimodal generation model capable of understanding and generating content across text, images, video, and audio, producing videos up to 15 seconds at 2K resolution with native stereo sound.
- H3 breaks away from siloed task-specific models by unifying generation across all modalities through a single pretraining paradigm, powered by key technologies including Contextual Omni Representation, H3-VAE, H3-Omni Transformer, and In-Context Regeneration, delivering over 4x efficiency gains and pricing at less than a third of mainstream competitors at 2K resolution.
- Model weights are set to be open-sourced in the coming days to support the developer community, with future development priorities including integrating M-series model capabilities, scaling model size, and improving visual fidelity and resolution.