Announcing Amazon Nova Multimodal Embeddings

Amazon Nova Multimodal Embeddings

Amazon Nova Multimodal Embeddings is now available, offering a unified embedding model for text, documents, images, video, and audio. This model simplifies cross-modal retrieval with leading accuracy, breaking down data silos and reducing complexity and costs.

Key features include:

• Unified Embedding Model: Supports multiple content types through a single model.
• Input Support: Handles up to 8K tokens and 30-second video/audio segments.
• Output Flexibility: Multiple embedding dimensions to balance accuracy and performance.
• API Options: Synchronous API for real-time applications and asynchronous API for large files.

Amazon Nova Multimodal Embeddings is available in US East (N. Virginia) in Amazon Bedrock.

What to do

• Read the AWS News blog for more information.
• Check the user guide to get started.
• Visit the Amazon Bedrock console to use Nova Multimodal Embeddings.

Source: AWS release notes

If you need further guidance on AWS, our experts are available at AWS@westloop.io. You may also reach us by submitting the Contact Us form.