Meta Movie Gen

Meta Movie Gen is an image to video tool. Meta's research suite of media foundation models for generating long HD video from text, editing existing footage with prompts, producing sound effects and soundtracks, and turning a personal photo into a video.

Meta Movie Gen is Meta's research program for a family of media foundation models that generate and edit video, produce synchronized audio, and create personalized clips from a single reference photo. The stated inputs are simple text prompts: a description of a scene, a requested change to existing footage, or a sound brief. Outputs include long high definition video at multiple aspect ratios, sound effects and full soundtracks, and edits that alter style, transitions or specific details of a clip. The system is presented as research first. The public page frames Movie Gen as a breakthrough benchmark effort, links to the underlying paper, and describes a creative industry feedback program in which Blumhouse selected filmmakers to try the suite before any public debut. Work from creators and directors is shown as evidence of what the models can do: transforming a dog into a baby dragon, turning a person into an alien, adding fog to a foreground, or building a selfie style clip of a person and their pet while keeping facial identity and motion intact. In practical terms the page suggests four capabilities: text to video generation, text driven video editing, personalized video from an uploaded photo, and audio generation that matches the tone, rhythm and style of footage. The pitch is aimed at filmmakers, storytellers, creators and researchers rather than casual editors, and the emphasis on directors and production companies signals where Meta sees adoption. Among alternatives, Movie Gen differs by treating video, sound and personalization as one connected research stack rather than separate consumer tools. The important caveat is availability: the page is a research showcase with a separate note that Meta AI now supports video editing, so anyone evaluating it for production should confirm which capabilities are actually reachable today and which remain demonstrations.