LTX Innovation & Technology Culture

LTX is building the most advanced open-source AI video technology in the world. LTX-2, released in January 2026, was the first complete open-source audio-video foundation model — generating fully synchronized video and audio within a single model. LTX-2.3, released March 2026, is the latest and most capable version, pushing visual quality and prompt adherence further. It remains the only open-source model of its kind in the Western world.

The open-source community reception has been significant. The models are available on HuggingFace, integrated into the Diffusers library, and actively extended by developers across ComfyUI, desktop apps, and custom pipelines worldwide. There's a dedicated Discord community for real-time development and collaboration — a direct signal of how quickly this technology has been adopted by builders globally.