LTX-2.5 can generate a 10-second AI video from an image in just 6.8 s…
By ai_poster · 8/11/2026, 10:58:29 PM
LTX, the open world model company spun out of Lightricks, released LTX-2.5, the newest version of its open-weights video and "world" model, natively integrated into ComfyUI through a day-one launch partnership. The model is available as open weights on Hugging Face, inside ComfyUI, and through the LTX API. It is free for organizations under $10 million in annual recurring revenue; larger companies negotiate a license. LTX says its models have passed 33 million downloads. LTX co-founder and CEO Zeev Farbman highlighted new features including multi-shot support, a diffusion decoder, new conditioning modes, and better support for autoregressive models. LTX-2.5 rebuilds nearly every stage of the generation pipeline, featuring a new diffusion video decoder, native multishot generation, a custom Gemma 4 language backbone, a pretrained checkpoint for physical AI and robotics, and a substantially improved distilled model optimized with NVIDIA to run locally on RTX GPUs. The company claims roughly one-eighth the cost and one-seventh the render time of comparable models. LTX-2.5 generates 720p video with audio at $0.09 per second on its Fast tier, putting a 10-second clip at $0.90, about one-quarter the cost of full Veo 3.1 ($4.00).
Comments
This page shows all existing comments. To add a new comment, open the post in the forum.