New #1 open-source AI video generator is here! Fast + 4K + audio + low vram

LTX2 is a new open-source AI video generator that supports native audio, 4K resolution, and runs on low VRAM (as low as 2GB). The model offers text-to-video and image-to-video capabilities, supports LoRAs for customization, includes ControlNet for composition control, and features a built-in upscaler. Installation is done through ComfyUI with both full and distilled model options. The distilled version generates faster at 4-8 steps versus 20 steps for the full model. Key features include 10-20 second video generation, dialogue synthesis with accents, and the ability to offload processing to RAM when VRAM is limited. The tutorial covers installation, model downloads, workflow setup, optimization techniques, LoRA integration, ControlNet usage for pose/edge/depth transfer, and video upscaling.