lightricks / ltx-video-0.9.7

DiT-based 13b video generation model, creating 30fps video

  • Public
  • 1.1K runs
  • GitHub
  • Weights
  • License
Iterate in playground

Run time and cost

This model costs approximately $0.14 to run on Replicate, or 7 runs per $1, but this varies depending on your inputs. It is also open source and you can run it on your own computer with Docker.

This model runs on Nvidia L40S GPU hardware. Predictions typically complete within 146 seconds.

Readme

LTX-Video-0.9.7 is the latest high-performance video generation model from Lightricks built on DiT. It can generate smooth, high-quality videos in real time—faster than you can watch them. At 1216×704 resolution and 30 frames per second, it produces realistic, dynamic scenes using text, images, or both.

You can control nearly every part of the output. Use a text prompt, or start with an image to animate it. Adjust the width, height, and number of frames to fit your needs. Fine-tune the motion and look of the video using parameters like negative prompt, guidance scale, and denoising steps. You can also set a custom frame rate and seed for repeatable results.