Video and image search in Amazon Bedrock Knowledge Base using Marengo 3.0

This article announces the general availability of TwelveLabs Marengo Embed 3.0 as an embedding model in Amazon Bedrock Knowledge Bases, enabling semantic search across video, audio, and image content.

• Marengo Embed 3.0 is a multimodal embedding model that encodes video, audio, images, and text into unified 512-dimensional vectors
• Managed Knowledge Bases automatically handle segmentation, frame sampling, transcription, and embedding generation without manual preprocessing
• Supports natural language queries like "show me penalty kicks" to find specific moments in hours of footage
• Available in US East (N. Virginia) and US West (N. California) regions with standard Bedrock model invocation pricing
• Walkthrough demonstrates creating a KB, syncing video data, running semantic searches, and integrating via Retrieve API
• Use cases span sports analytics, media asset management, security surveillance, education, and retail product demos

Marengo Embed 3.0 in Amazon Bedrock Knowledge Bases eliminates the need for complex custom pipelines, making multimodal video and image search fully managed and accessible.