Genmo launches AI video generation model Mochi 1 to compete with Sora, Runway, and others

Genmo has unveiled its latest innovation, Mochi 1, a powerful and open-source video generation model. This release marks Genmo's first public open-source offering, licensed under Apache 2.0, allowing for both individual and commercial use.

Mochi 1 is designed to generate video content with realistic motion that adheres to the laws of physics, offering users detailed control over characters, settings, and actions. It aligns video output closely with textual prompts, ensuring consistent and fluid human action and expression. Preliminary evaluations highlight its high-fidelity motion and strong prompt adherence, narrowing the gap between open and closed video generation systems.

Mochi 1 is now accessible on Genmo's Playground and on the Hugging Face platform, providing users with the opportunity to explore its capabilities.