aws AWS What's New ·

Amazon Bedrock Knowledge Base adds TwelveLabs Marengo 3.0 for multimodal embeddings

aiawsgaengineerretailmediaeducationaws-s3aws-bedrock
feature announcement

Amazon Bedrock Managed Knowledge Base now integrates TwelveLabs Marengo 3.0, expanding its capabilities to generate multimodal embeddings for video, audio, and image content. This enhancement allows for advanced media search by directly encoding visual scenes, speech, and video cues, capturing meaning beyond text transcription. Customers can upload media assets from sources like Amazon S3 and perform natural language searches without managing infrastructure. The model produces compact 512-dimensional vectors and provides segment start/end times, enabling applications to jump to specific moments in a video for various use cases, including sports analytics, media, education, and retail.

Features (1)
  • Multimodal Embeddings with TwelveLabs Marengo 3.0

    Amazon Bedrock Managed Knowledge Base now supports TwelveLabs Marengo 3.0, enabling the creation of multimodal embeddings for video, audio, and image content. This model encodes visual scenes, speech, and video cues directly into embeddings, going beyond text transcription to capture richer meaning for retrieval. It supports natural language search, delivers state-of-the-art retrieval accuracy with 512-dimensional vectors, and provides segment start and end times.

Read the original announcement →

https://aws.amazon.com/about-aws/whats-new/2026/09/amazon-bedrock-managed-knowledge-base-multimodal-embeddings-twelvelabs-marengo/

Related releases