Video models are getting better all the time but even the fastest ones can take a bit of time to generate videos from text and images. Lucy-14B aims to change that. It generates photorealistic videos and is up to 7x faster than WANv2.2-14B. It can generate 5s videos in 6.5 seconds. Best of all, you don’t have to make a compromise on quality.
Yes. You can have it all.
Compared to WANv2.2-14B:
– 7× faster
– 5s video in ~6.5s
– Same quality pic.twitter.com/yVZ3tt0xHc— Decart (@DecartAI) September 10, 2025
This model is currently available on fal. You will just have to enter your text prompt, upload your image, and the rest is easy. Each second of video will cost $0.08.
[HT]
