>
May, 5th, 2025: New model 13B v0.9.7: Release a new 13B model ltxv-13b-0.9.7-dev Release a new quantized model ltxv-13b-0.9.7-dev-fp8 for faster inference with less VRam Release a new upscalers ltxv-temporal-upscaler-0.9.7 ltxv-spatial-upscaler-0.9.7 Breakthrough prompt adherence and physical understanding. New Pipeline for multi-scale video rendering for fast and high quality results
Training Code Accessibilitycommercial license is required for entities with $10M+ revenues https://huggingface.co/Lightricks/LTX-Video-0.9.7-dev Apache 2.0 for inference and fine-tuning code: https://github.com/Lightricks/LTX-Video
Parameters13000000000
Notes: 13B