TwelveLabs

A multimodal AI platform that searches, analyzes, and understands video content across vision, audio, and language, solving the problem of extracting insights from large video libraries at scale. It sells to organizations working with video at scale, including media, sports, advertising, government, and security teams. The platform is positioned as an infrastructure pipeline that indexes video at 60x real-time speed, turning passive footage into a strategic asset with models like Marengo for embedding and Pegasus for reasoning.

Key features

• Ingest multimodal data through a single pipeline
• Index an hour of video in a minute
• Search video libraries using natural language
• Locate specific actions, scenes, dialogue, emotions
• Segment content
• Ensure compliance
• Create highlights
• Generate insights
• Marengo multimodal embedding model
• Pegasus video language model
• Supports 47 languages
• Encrypted data at rest and in transit
• Role-based access control
• Multi-factor authentication
• Audit and access logging
• SOC 2 Type II certified