Meta released its video generation model "Muse Video" in preview on July 7, 2026, debuting at No. 3 in the community-voted Text-to-Video Arena with an Elo score of 1459. Muse Video belongs to "Muse," the first family of media generation models developed by Meta Superintelligence Labs (MSL), and shares the same pretraining base as the image model "Muse Image," which was formally released at the same time. The score topped Alibaba's HappyHorse 1.0 by 30 points and ranked ahead of Grok Imagine, Sora 2 Pro, and Google Veo-3.1. According to the announcement, Muse Video features high visual fidelity and native audio support, and is set to be offered soon to creators and within Meta AI.
Continue reading
The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in✓ Signed in — this article isn’t included in your current plan.