Cohere’s Embed v4 multimodal embeddings model now available on Amazon Bedrock

Amazon Bedrock Updates

Amazon Bedrock now offers Cohere Embed v4, a state-of-the-art multimodal embedding model for text, images, and complex business documents. This model enhances search and retrieval capabilities for AI applications.

Key features include:

• Multimodal Processing: Handles documents with tables, graphs, diagrams, code snippets, and handwritten notes.
• Language Support: Supports over 100 languages, including Arabic, English, French, Japanese, and Korean.
• Industry Fine-Tuning: Optimized for finance, healthcare, and manufacturing documents.

What to do

• Access Embed v4 in US East (N. Virginia), Europe (Ireland), and Asia Pacific (Seoul).
• Use cross-region inference from select public AWS Regions.
• Request model access via the Amazon Bedrock console.

Source: AWS release notes