Generate LTX-2 videos with native 4K & synchronized audio
LTX-2.3 is Lightricks' newest open-source AI video model — true 4K up to 50fps, audio and video generated together, text-to-video and image-to-video in 6–10 second clips.
What's new in LTX-2
A generational leap over previous LTX Video models — built for creators, marketing teams and professional production.
Synchronized audio, natively
LTX-2 is the first fully open model to generate video and matching sound together — ambient audio, music, effects and accurate lip-sync — all in a single inference pass.
True 4K at up to 50fps
Render production-grade clips at native 4K (2160p) with sharp textures and clean motion, or iterate fast at 1080p. You choose the fidelity per generation.
Text-to-Video & Image-to-Video
Start from a text prompt, or bring a still image to life. LTX-2 animates your image while keeping its composition and style coherent.
Fast, Pro & Ultra modes
Fast mode for rapid previews, Pro for balanced quality, and Ultra for maximum-fidelity 4K with audio — pick the trade-off that fits your workflow.
Fast, Pro & Ultra
Choose the quality tier that matches your workflow — every mode runs the latest LTX-2 model.
Quickest turnaround for previews and iteration. Powered by LTX-2 Fast.
- Lowest credit cost
- 1080p output
- Great for drafting
Balanced fidelity and speed for everyday content and marketing. Powered by LTX-2 Pro.
- Up to 4K
- Strong detail & motion
- Recommended default
Maximum fidelity for cinematic and professional use — forced 4K with synchronized audio.
- Native 4K (2160p)
- Synchronized audio on
- Best for final renders
LTX-2 FAQ
Try the latest LTX-2 model now
Turn text or an image into a 4K video with synchronized audio in seconds.