Muse Image and Muse Video, the first media generation models developed by Meta (twitter.com)

Meta Superintelligence Labs has unveiled Muse Image and Muse Video, its first media generation models that promise to transform how content is created and experienced. Muse Image is an advanced image generation model capable of accurately following user instructions, making precise edits, and composing visuals from multiple references, all while integrating social context from Instagram. Additionally, it supports agentic tool use, enhancing user interactivity with the generated images. Users can access Muse Image through the Meta AI app, on the web, and within Instagram Stories and WhatsApp, initially available in select countries. Muse Video, which shares the same foundational pretraining as Muse Image, aims to deliver high visual fidelity and now includes support for native audio, making it a powerful tool for video content creation. The introduction of these models is significant for the AI/ML community as they represent a leap forward in generative AI capabilities, particularly in blending visual and auditory media. This advancement positions Meta at the forefront of media generation technology, enabling creators to produce engaging content more efficiently and dynamically.