What is Voyage AI used for?

Voyage AI, now part of MongoDB, provides embedding and reranking models that improve search and retrieval-augmented generation over unstructured data. It is a good fit for improving retrieval quality in RAG systems, semantic search over unstructured data, and reducing vector storage and search costs.

Yes. Voyage AI has a free plan.

Is Voyage AI open source?

No. Voyage AI is proprietary (closed-source) software. In the AI Infrastructure category, open-source options include Supavec, Superduper and Instill Core.

Can I self-host Voyage AI?

Yes. Although Voyage AI is closed source, it can be self-hosted on your own servers, and the vendor also offers a hosted version.

What are some alternatives to Voyage AI?

Voyage AI competes with Cohere, Jina AI and OpenAI API Platform.

Voyage AI: features, pricing and alternatives

About Voyage AI

Voyage AI builds embedding models and rerankers used for search and retrieval, especially in retrieval-augmented generation systems. The company is now presented as part of MongoDB and its Intelligent Data Platform, and it targets teams that need relevant results from unstructured data.

The vendor lists several design goals: high retrieval accuracy on an enterprise-oriented benchmark, lower-dimensional vectors that cut vector search and storage costs, lower latency, long context lengths of up to 32K tokens, and compatibility with any vector database or language model. Embeddings convert text into vectors and rerankers reorder results to improve the quality of what a model sees.

Models are available as a hosted service with pay-as-you-go pricing, through major cloud providers, or deployed in a customer's own VPC. The company also takes inquiries for custom or on-premise deployments and for model licensing.

Key features

Text embedding models for semantic search
Rerankers that reorder retrieved results
Long context support up to 32K tokens
Shorter vectors to reduce storage cost
Works with any vector database or LLM
Cloud, in-VPC and on-premise deployment

Good fit for

→Improving retrieval quality in RAG systems
→Semantic search over unstructured data
→Reducing vector storage and search costs

Tags: embeddings; rerank; rag; semantic-search; vector-search; mongodb; ai-infrastructure

Voyage AI: questions and answers

What is Voyage AI used for?: Voyage AI, now part of MongoDB, provides embedding and reranking models that improve search and retrieval-augmented generation over unstructured data. It is a good fit for improving retrieval quality in RAG systems, semantic search over unstructured data, and reducing vector storage and search costs.
Is Voyage AI free?: Yes. Voyage AI has a free plan.
Is Voyage AI open source?: No. Voyage AI is proprietary (closed-source) software. In the AI Infrastructure category, open-source options include Supavec, Superduper and Instill Core.
Can I self-host Voyage AI?: Yes. Although Voyage AI is closed source, it can be self-hosted on your own servers, and the vendor also offers a hosted version.
What are some alternatives to Voyage AI?: Voyage AI competes with Cohere, Jina AI and OpenAI API Platform.