UnstableML

UnstableML hosts a collection of some of the best SOTA video diffusion models with seamlessly integrated synchronized audio to deliver an unparalleled HD videos for content creators. We also host our custom video-to-video model, Dream, designed specifically for video editing. With a simple prompt Dream can edit, replace, transform or generate similar videos based on your input video. We also feature our custom enhancer model designed to add intricate detail, improve dexterity, and eliminate unknown artifacts from videos, while elevating them to stunning 2K resolution and up to 120 FPS.