LTXV-13B

About

A high-performance Diffusion Transformer (DiT) designed for realistic video synthesis from text or image prompts 🚀

Key Features

• 13-billion parameter DiT architecture 🧠
• High-resolution text-to-video and image-to-video 🎥
• Superior temporal consistency and motion fluidity 🎞️

Pros

• Exceptional motion realism and accuracy 🏃
• Very fast inference for its size ⚡
• Fully open weights for local deployment 💻

Cons

• High VRAM requirements (24GB+) 🔋
• Limited to short clip generation ⏱️