Captions changes its name to Mirage to expand its AI video technology

Captions, the platform used by over 20 million content creators and businesses, has announced its name change to Mirage as part of an expansion strategy toward developing more advanced artificial intelligence models for short-form video creation. The company maintains its original application under the Captions brand but expands its vision toward research of foundation models that can generate complete audiovisual content from a single prompt.

According to the company, short-form video currently represents the most popular and fastest-growing communication medium, especially on platforms like TikTok and Instagram Reels. However, unlike text or image generation through AI, video creation still requires significant manual editing work before being ready for publication. Mirage identifies this gap as a market opportunity to develop technology that produces directly usable audiovisual content.

The company has developed a proprietary model trained specifically to optimize pacing, framing, and attention dynamics characteristic of the short format. This approach allows generated videos to maintain the visual appearance and feel typical of social networks, automatically adapting to the engagement standards of these platforms. Mirage's multimodal model can convert voice and language into photorealistic video, synchronizing speech with natural lip movements, eye contact, and emotional expressions.

The company's research spans real-world generative modeling, multimodal reasoning, dataset design, audio-video quality evaluation, and large-scale training. Its stated goal is to reduce the distance between idea and audience to zero, allowing creators and businesses to develop their full creative and commercial potential through end-to-end foundation models that natively process from script to audio-video.

Key Points

• Captions changes its name to Mirage to reflect its new expanded vision toward artificial intelligence foundation models for video.
• The company has over 20 million users among content creators and businesses using its tools.
• Mirage identifies a market gap: while text and images are generated with AI through a single instruction, video still requires significant manual editing.
• The company has developed a proprietary model trained specifically to optimize pacing, framing, and attention dynamics of the short format.
• Its multimodal technology converts voice and language into photorealistic video with automatic synchronization of lip movements and expressions.
• Generated videos are optimized for platforms like TikTok and Instagram Reels, automatically adapting to their engagement standards.
• The original Captions application remains active as a central part of the new business strategy under the Mirage brand.