Show HN: MiniMax H3 – Turn text and images into AI video clips (minimax3.com)

MiniMax has launched its latest tool, MiniMax H3, which allows users to generate AI-driven video clips using text and image inputs. This new capability supports video outputs ranging from 4 to 15 seconds in 768P and 2K resolutions, alongside native stereo audio. Users can craft detailed prompts that describe complete shots while incorporating multimodal references, such as images, video, and audio files, to guide the visual and auditory elements of the generated scenes. The tool emphasizes the significance of clarity in prompts, enabling focused edits on characters, objects, and dialogue. This development is noteworthy for the AI/ML community as it merges multiple forms of media into a cohesive video creation process, enhancing the capabilities available to content creators, marketers, and filmmakers. The official API allows developers to integrate MiniMax H3 into their workflows, supporting rapid prototyping and testing with visual references. As generative AI continues to evolve, tools like MiniMax H3 signal a shift toward more sophisticated, user-friendly content production solutions, bridging the gap between creative ideation and final output. Users can start with free credits before deciding on a subscription model, making it accessible for individuals and teams alike.