We’re excited to introduce LTX-2, the latest video generation model supported in ComfyUI as Partner Nodes (API Nodes).
Available now:
Text-to-Video and Image-to-Video Models, both with synchronized audio
Models: LTX-2 (Pro), LTX-2 (Fast)
Up to 4K resolution at 50 FPS
Duration: 6 / 8 / 10 Seconds
Model Highlights
Efficient Generation: Streamlined inference enables multi-scene iterations in seconds.
Multimodel Generation: Generations with synchronized audio.
Intelligent Motion Understanding: Generates smooth temporal coherence and realistic camera movement.
Enhanced Texture & Surface Detail: New diffusion backbone trained for physical material accuracy.
Get Started
Update ComfyUI to the latest version.
Open the LTX-2 template workflows in the Template Library.
You can also search for “LTXV Text To Video” or “LTXV Image To Video” Nodes and build your own workflows!
More Examples
As always, enjoy creating!