Emu Video by Meta

Emu Video by Meta is a state-of-the-art AI tool for text-to-video generation, leveraging diffusion models to create videos from text prompts. The process is factorized into two stages: generating an image from the text, then generating a high-quality, photorealistic video based on both the text and the generated image. Designed for researchers and creators wanting to produce 4-second, 512px, 16fps videos, Emu Video outperforms previous models in video quality and adherence to prompts.