Where Hyperflow H3 is strongest in Image to Video: only Utopai X scores higher on Multi-Scene & Narrative, and Spatio-Temporal Consistency is where it improves most on MiniMax H3 (768p)
Capabilities are the model skills a prompt is written to test. In AA-Video-I2V v2.1, every prompt pairs a starting image with a text prompt written for that image.
Against MiniMax H3 (768p), the model it builds on, Hyperflow H3's largest gains are in Spatio-Temporal Consistency and Audio Synchronization. Human Anatomy and Dialogue & Lip Sync are where it trails the leaders most.
➤ Multi-Scene & Narrative covers multi-event stories, shot transitions, montage, and identity consistency across shots.
➤ Spatio-Temporal Consistency covers spatial coherence, time lapses, occlusion and re-entry, and single-shot stories.
➤ Human Anatomy, where Hyperflow H3 trails the leaders most, covers gait and locomotion, fine motor movement, facial expression, acrobatic motion, object grasping, and human interaction.