This MiniMax H3 ComfyUI workflow is designed for audio-driven long-form videos and digital human performances. Supply a character reference image, an audio track, and prompts for each segment. The workflow trims audio into specified intervals, generates video segments, and uses the final frame of each segment to guide the next, helping maintain visual continuity before automatically combining the video and audio. Suitable for talking avatars, character singing performances, music videos, and sequential storytelling. Includes adjustable aspect ratio, resolution, segment duration, and optional RIFE frame interpolation.