LTX-2 is an open-source AI video generation model hosted on Vidthis that produces native 4K, up to 50 fps, audiovisual content with perfectly synchronized sound. It supports both text-to-video and image-to-video modes, offers durations from 5 to 20 seconds, and provides granular camera, motion, and style controls via LoRA adapters. LTX-2 targets production workflows with Fast/Pro/Ultra modes, reproducible seeds, and the ability to customize or deploy weights locally for advanced fine-tuning.
LTX-2 Core Features
Open-source LTX-2 model weights (GitHub/Hugging Face)
Explore Free génération vidéo avec son synchronisé Tools and Resources
Unlock the potential of free génération vidéo avec son synchronisé tools. Simplify workflows, enhance efficiency, and achieve results—all without spending a dime.