TwelveLabs is a video-native AI platform that enables enterprises to analyze, search, and generate insights from video content using multimodal AI models. Its technology, including the Marengo and Pegasus model families, is designed to understand video at a deep semantic level, setting it apart from traditional text-based or single-modality approaches. The company has raised over $100M in Series B funding from investors including NEA and NVIDIA, underscoring its credibility and momentum in the AI video intelligence space. Its platform is used across media, advertising, and automotive industries for tasks such as content moderation, video search, and automated workflow generation.
TwelveLabs outlines an architectural blueprint for streaming publishers to operationalize in-house video intelligence. By integrating multimodal AI models like Pegasus 1.5 and Marengo 3.0 with ad-serving infrastructure such as FreeWheel and AWS Elemental MediaTailor, the workflow targets advanced, brand-safe contextual targeting for VOD, FAST, and live sports.
Video-AI startup TwelveLabs has integrated its Marengo video understanding model natively into the Snowflake AI Data Cloud. This allows media, entertainment, and advertising teams to process video files and generate vector embeddings directly within their secure Snowflake environments, enabling analytics on unstructured video data. The integration aims to support advanced metadata queries, brand suitability scoring, and content curation for platforms like Warner Bros. Discovery, Disney, and Paramount.
TwelveLabs has announced Marengo 3.0, a new multimodal embedding model designed for video retrieval. This iteration supports advanced features including composed queries, multilingual search capabilities, and the processing of long-form video content.
Twelve Labs introduced Pegasus 1.5, an update to its AI model that transforms video into structured, time-based metadata. This version utilizes schema-driven segmentation, custom evaluation metrics, and reinforcement learning to align with real-world video workflows.
