Hailuo 2.3 Text-to-Video Capability Illustrated: Parameters and Scenario Guide
A short film is used to illustrate the expressive direction of the MiniMax Hailuo 2.3 text-to-video path: pure text prompts directly drive the finished clip, with a focus on motion smoothness, anime-style stability, the realism of full-body movement and subtle facial expressions, and the reliability of finished output in commercial and product scenarios. This work uses only the three inputs declared in the author's pipeline - prompt, duration (6 or 10 seconds), and prompt_optimizer (enabled by default) - and does not involve first/last frames or multi-image references, consistent with the official documentation's boundary that "first and last frames are not supported, and generation is based only on a single first frame/single prompt." In prompt writing, the subject, scene, camera language, and pacing are all clearly laid out within a single block of text, so that intent stays stable even after prompt_optimizer is enabled. It is suitable for anime short films, dance and full-body action performances, live-action-feel shots requiring realistic facial emotion, and rapid test shoots and script visualization for product promo videos; compared with Hailuo 2.3-Fast, this model covers the text-to-video path and is well suited for direction validation before formal production.
Style Text-to-Video · Anime style · character-motion
Model Hailuo 2.3
Prompt
A continuous cinematic shot, in anime style: a young dancer in a loose white jacket walks into a rain-soaked, neon-lit alley at night, first performing a slow full-body spin, then flowing seamlessly into a rapid footwork sequence; the camera tracks from a low angle, moving in a half-orbit with slight handheld shake, always keeping the dancer's full body in frame so the limb movements are clearly legible; rim light from pink and cyan signs outlines the edge of the jacket, shallow depth of field, the wet asphalt reflecting a glossy sheen; the shot ends with a slow push-in to a medium close-up as the dancer turns toward the camera, a subtle shift in expression, visible breath, and the hint of a smile at the corner of the mouth; consistent character design, stable style rendering, natural motion, 24fps cinematic motion blur, 6 seconds.
Copy the full prompt and adapt it for your own generation.
Tags