LTX-2 is an open-source multimodal AI model developed by Lightricks that generates high-quality video clips with perfectly synchronized audio in a single generation process. It produces native 4K resolution videos at up to 50 FPS with coherent motion and audio alignment for up to 20 seconds, all while running efficiently on consumer-grade NVIDIA GPUs without requiring cloud services.
Key benefits include:
- One-Shot Audio-Video Sync: Generates video and perfectly synchronized audio (dialogue, sound effects, music) in a single diffusion process, eliminating post-production dubbing and alignment
- Professional 4K Quality: Outputs cinematic-quality video at up to 4096×2160 resolution and 50 FPS with consistent lighting and reduced flicker
- Extended Coherent Duration: Creates up to 20 seconds of continuous video with maintained visual consistency and narrative coherence
- Local GPU Deployment: Optimized for NVIDIA RTX GPUs with FP4/FP8 weights, enabling 4K generation on consumer hardware
- Advanced Creative Control: Supports text prompts, image inputs, sketches, and ComfyUI workflows with Pro/Ultra quality modes
Perfect for independent filmmakers, content creators, and studios producing professional short-form content who need synchronized audio-video generation without cloud dependencies.
