Voyage AI Provider
The Voyage AI Provider brings powerful embedding and reranking capabilities to the Vercel AI SDK. Generate high-quality text and multimodal embeddings, or improve your search results with state-of-the-art reranking models.Why use Voyage AI Provider?
Voyage AI offers specialized embedding models optimized for different use cases:- Domain-specific models - Choose from specialized models for code, finance, law, and multilingual content
- Multimodal capabilities - Generate embeddings from text, images, or both combined
- Flexible dimensions - Select output dimensions (256, 512, 1024, or 2048) to balance quality and performance
- Advanced reranking - Improve search relevance with dedicated reranking models
- Large context windows - Process up to 32,000 tokens in a single request
Key features
Text embeddings
Generate embeddings for text with specialized models for different domains like code, finance, and law
Image embeddings
Create embeddings from images for visual search and similarity matching
Multimodal embeddings
Combine text and images into unified embeddings for rich content understanding
Reranking
Reorder search results by relevance to improve search quality
Get started
Quickstart
Generate your first embeddings in minutes
Installation
Install the provider with your preferred package manager
Text embeddings
Learn how to generate text embeddings
API reference
Explore available models and configuration options
Available models
Voyage AI offers multiple embedding model families:- voyage-3.5 series - Latest models with flexible output dimensions
- voyage-code-3 - Optimized for code understanding and search
- voyage-finance-2 - Specialized for financial documents and analysis
- voyage-law-2 - Tuned for legal text processing
- voyage-multilingual-2 - Supports 100+ languages
- voyage-multimodal-3 - Handles text and images together