Video AI
Twelve Labs launches Marengo Embed 3.0 video model on Amazon Bedrock
Marengo becomes the first video-specific embedding model integrated into Bedrock's fully managed knowledge base service, enabling natural language visual search.
Original cover art, generated for this story. THE VISSION does not republish third-party press imagery.
The short version
- Twelve Labs has announced the general availability of its Marengo Embed 3.0 model as a managed service within Amazon Bedrock.
- The model is the first video-specific embedding tool on Bedrock, encoding visual scenes, speech, and audio into unified 512-dimensional vectors.
- The integration allows enterprise AWS customers to build precise semantic search across video libraries using natural language queries.
Marengo becomes the first video-specific embedding model integrated into Bedrock's fully managed knowledge base service, enabling natural language visual search.