TwelveLabs Inc. raises $100M in Series B

TwelveLabs Inc., a developer of generative AI models focused on video understanding, has secured $100 million in a Series B funding round. This capital injection, critical for its advancement of AI capabilities in video content analysis, underscores the company's endeavor to enhance artificial intelligence to comprehend videos similarly to how humans do. The round was spearheaded by NEA and NAVER Ventures, with significant contributions from Amazon, Radical Ventures, Korea Investment Partners, Index Ventures, Quadrille Capital, and Red Bull Ventures, bringing the total equity raised by TwelveLabs to over $207 million.

Co-founder and CEO Jae Lee emphasized the company's strategic focus on video as a primary data substrate, diverging from traditional language models: "The substrate of machine intelligence is recorded reality in motion, not language," he noted. TwelveLabs has pioneered the development of foundational AI models designed specifically to interpret video. Their flagship model suite, including Marengo 3.0, converts video, text, and audio into machine-readable data formats facilitating advanced comprehension and retrieval capabilities. Another model, Pegasus, complements these capabilities by converting video into structured data, akin to how language models process documents.

The strategic rationale for this funding aligns with TwelveLabs' mission to transform video perception across various sectors, such as security, advertising, sports, and automotive industries. These industries generate vast amounts of video data, which require sophisticated AI for effective analysis and operationalization. By constructing solutions that allow AI to natively understand and reason across video content, TwelveLabs aims to establish a new paradigm in video intelligence.

In terms of market dynamics, TwelveLabs' advancements place it at the forefront of the AI video modeling sector, challenging current limitations in video data processing. Today's large language models struggle with video content, requiring segmentation into discrete frames for analysis. TwelveLabs' models, however, are engineered to maintain a comprehensive understanding over time, developing intelligence with each interaction. This positions the company as a key player amid growing demand for advanced AI solutions capable of handling complex multimedia inputs.

Looking ahead, TwelveLabs is set to deepen its collaboration with Amazon Web Services (AWS), with the recent funding round facilitating a multiyear commitment to optimize its video inference workloads on AWS infrastructure. Future frontier models are slated for initial deployment on AWS, indicating a strong pathway for technological integration and expanded service offerings within the cloud ecosystem.