Meta’s AI model, Muse Video, has officially entered the Video Arena at the third position, scoring 1459 and surpassing Alibaba’s HappyHorse 1.0 by 30 points. This achievement marks a significant milestone for Meta Superintelligence Labs, which developed both Muse Video and the advanced image model Muse Image. Both models feature strong capabilities, with Muse Video offering exceptional visual fidelity and native audio support. Muse Image is currently available through the Meta AI app, Instagram Stories, and WhatsApp in selected countries, showcasing Meta’s commitment to advancing media generation technologies.
Meta: Meta is a technology company focused on social connectivity and artificial intelligence development. Its Meta Superintelligence Labs created the Muse suite of media generation models. Muse Video, the company’s new video model, has entered the Video Arena at the number three position.
AI at Meta: AI at Meta serves as the primary channel for announcing Meta’s artificial intelligence projects and model releases. The account shared details on the launch of Muse Image and Muse Video from Meta Superintelligence Labs. It highlighted Muse Video’s strong performance in the Text-to-Video Arena.
Muse Image: Muse Image is Meta’s latest image generation model, designed for precise instruction following, editing, and multi-reference composition. It incorporates Instagram social context and agentic tool use capabilities. The model integrates with Muse Spark and is rolling out in the Meta AI app along with Instagram Stories and WhatsApp.
Muse Video: Muse Video is a media generation model developed by Meta Superintelligence Labs as a preview release. It shares the same pretraining foundation as Muse Image while adding native audio support and high visual fidelity. The model achieved a top-three ranking in the Video Arena shortly after introduction.
Availability: Muse Image is being made available through the Meta AI app, Instagram Stories, and WhatsApp in initial regions.
AI Development: Meta Superintelligence Labs produced its first media generation models in the form of Muse Image and Muse Video.
Model Capabilities: Muse Video delivers native audio support alongside strong visual quality as part of its core design.