Voyage Large 2
Voyage Large 2 is Voyage AI's higher-capacity embedding model, offering stronger semantic fidelity than Voyage 2 by using a larger underlying architecture. It is designed for scenarios where embedding quality is the priority, such as precision-critical retrieval over long or technical documents where subtle semantic distinctions need to be preserved.
Voyage positions this model for demanding enterprise retrieval workloads — legal document review, scientific literature search, and complex multi-hop RAG — where the marginal cost of a larger model is justified by improved recall and ranking quality. It shares the same retrieval-first training philosophy as Voyage 2 but with greater representational capacity.
Key Features
Higher representational capacity than Voyage 2 for nuanced semantics
Excels on long-form and domain-specific document retrieval
Improved performance on re-ranking and passage-level matching
Well-suited for multi-hop and complex question answering retrieval
Compatible with vector databases such as Pinecone, Weaviate, and Qdrant
Ideal Use Cases
High-precision legal and compliance document retrieval
Scientific literature search and citation recommendation
Enterprise knowledge base search requiring nuanced ranking
Multi-hop RAG pipelines over large heterogeneous corpora
Semantic deduplication in publishing or research workflows
Example Prompts for Voyage Large 2
Technical Specifications
| Provider | Voyage |
| Category | Embedding |
| Modality | Text -> Vector |
Frequently Asked Questions
Try Voyage Large 2 now
Start using Voyage Large 2 instantly — 100 free credits, no credit card required. Access 750+ AI models through one platform.