Showcases

From Text to Visuals: Hailuo 2.3 Text-to-Video Example

This entry uses the text-to-video path of Hailuo 2.3 to demonstrate the transformation from "text to visuals": no reference image is provided; generation is driven solely by an English prompt containing the subject, action, camera, and lighting, testing how well the model understands and realizes the storyboard description. The prompt is organized in storyboard style—first establishing the character and scene, then describing full-body action and camera movement, and finally constraining lighting and style—making it easy for developers to compare against the submit input and observe which textual information is consistently reproduced. Duration is set to 6 seconds, keeping the default prompt optimization toggle. Applicable scenarios: rapid visualization of creative scripts, storyboard concept test shoots, and baseline samples for evaluating the controllability of text-to-video.

Style Text-to-Video · Storyboard · Cinematic

Model Hailuo 2.3

Try it yourself

Prompt

Try this model

Cinematic single-shot concept: a lone courier in a worn olive jacket steps out of a rain-slicked alley and strides quickly toward the camera, one hand shielding a small parcel, the other brushing along the damp wall. The camera tracks backward at walking speed, at eye level, slightly low, keeping the subject centered; a shallow depth of field softly blurs the neon signs and the steam rising from street vents behind. Raindrops fall as visible thin streaks, and puddles reflect red and cyan light. The jacket fabric and hair have natural motion, and subtle expression changes appear on the subject's face as they glance back. Cool teal and amber color grade, soft ambient practical light, cinematic contrast, 24fps cadence, no cuts, continuous steady movement.

Copy the full prompt and adapt it for your own generation.