This all-in-one ComfyUI workflow combines MiniMax H3, SelfLift, and VOSR2 for video creation, bringing together local Qwen prompt expansion, multiple image and audio references, audio-video generation, and optional post-processing. SelfLift starts sampling at a lower resolution before switching to high-resolution refinement, with adjustable transition and upscaling correction settings. Optional VOSR2 and RTX super-resolution stages, followed by RIFE frame interpolation, connect generation, upscaling, and interpolation in one workflow. Customize the aspect ratio, resolution, duration, and LoRAs for character animation, narrative shorts, and multi-reference video projects.