LAS VEGAS—Amazon Web Services (AWS) has partnered with TwelveLabs to introduce advanced multimodal foundation models, named Marengo and Pegasus, within Amazon Bedrock. This initiative marks a significant milestone as AWS becomes the first cloud provider to feature models by TwelveLabs.

TwelveLabs aims to revolutionize video search capabilities, capitalizing on the vast amounts of previously unsearchable video data. With video content accounting for nearly 80% of the world’s data, its integration into Amazon Bedrock is anticipated to streamline video archiving and accessibility for creators and library owners.

Amazon Bedrock is a fully managed service that provides developers access to high-performing models via a single API. This collaboration will drive transformative changes in how enterprises analyze, search, and generate insights from video content, bolstered by AWS’s secure and robust infrastructure.

Jae Lee, co-founder and CEO of TwelveLabs, emphasized the potential of this technology: ‘Our models will empower enterprises to efficiently search video libraries—whether the content was produced yesterday or a decade ago—in under a second.’ This capability aims to unlock numerous opportunities for innovative applications.

The collaboration promises invaluable benefits across multiple sectors, including media, entertainment, and advertising. For instance, video production teams can streamline workflows related to video content management and create targeted content efficiently.

• Natural language video search to pinpoint specific content moments.
• Deep video understanding without pre-defined labeling.
• Multimodal AI processing, addressing visual, audio, and text elements.
• Temporal intelligence connecting related events through time.
• Enterprise solutions that scale extensive video libraries.

As this integration unfolds, AWS customers will leverage TwelveLabs’ models to develop generative AI applications, enhancing their overall data management strategies while maintaining enterprise-grade security and control over their information.

Samira Panah Bakhtiar, general manager of Media & Entertainment at AWS, stated: ‘Natural language semantic search is a strategic unlock for our entertainment customers, enabling them to derive maximum value from their intellectual property.’ This collaboration signifies a pivotal shift in harnessing video data to humanize user experiences and streamline operational efficiency.

This announcement builds on enhancing the ongoing collaboration between TwelveLabs and AWS, making advanced AI video understanding models more accessible to enterprises globally.