Mirage is a New York City-based company developing foundation models and products aimed at transforming short-form video production. Its core work involves building generative models that convert voice and language input into photorealistic video output.
The company's technical focus spans several domains, including generative AI, multimodal AI, computer vision, speech synthesis, and video generation. Its primary product is a multimodal foundation model designed to match speech with natural lip movement, eye contact, and emotional expression, producing A-roll footage intended to look and feel realistic.
Mirage operates within the content creation and video production verticals. The team comprises researchers, engineers, and creators working at the intersection of language, speech, and visual media.
