fal-ai/ltx-2.3-quality/colorization
Colorize high-quality video using LTX-2.3
Inference
Commercial use
Prompt examples
Examples are generated using the Ltx 2.3 Quality. You can customize them by clicking on the "Playground" button.
Reference shows a bustling late 19th or early 20th-century city street rendered in high-contrast monochrome. Pedestrians in period clothing, horse-drawn carriages, and early motor vehicles move down a cobblestone avenue lined with multi-story brick and stone buildings under a soft, overcast daylight that casts gentle shadows.
Edited shows the same scene with natural, authentic colors restored. COLORIZE the historical street scene with realistic, muted tones. The buildings feature weathered red and brown brickwork alongside ornate cream-colored limestone facades and canvas awnings in soft beige and off-white. The cobblestone street reflects a subtle mix of earthy greys and charcoal. Pedestrians wear dark woolen coats in deep charcoal, navy, and brown, contrasted by a young boy in a muted linen shirt and dark trousers. The horse-drawn carriage in the foreground is deep polished chestnut wood with a dark horse, while the early electric streetcar in the background displays a classic vintage green and cream livery. The overall lighting is soft and natural, maintaining a rich, cinematic 1990s film stock grading with true-to-life skin tones and balanced color saturation. Subject identity, framing, and motion geometry are identical to the reference; only color information differs between reference and edited.
num_inference_steps15
guidance_scale1