MODEL · EMBEDDINGS
voyage-3: high-quality retrieval embeddings with a 32K token context.
voyage-3 is a text embedding model from Voyage AI built for retrieval accuracy. With a 32K token context window, it converts text inputs into dense vector representations suited to semantic search, retrieval-augmented generation (RAG) pipelines, and grounded AI applications. Access it inside AresGen without managing separate Voyage AI API credentials.
Strengths
What voyage-3 brings to your workflows
- Text embedding generation converts text inputs into dense vector representations — provide documents, queries, or passages and voyage-3 produces embeddings that capture semantic meaning for downstream retrieval, similarity, and classification tasks.
- 32K token context window allows embedding long documents, extended passages, and multi-paragraph content in a single request without chunking overhead that can affect retrieval quality.
- Retrieval-optimised embeddings are suited to semantic search and RAG pipelines where retrieval accuracy determines the quality of grounded AI responses — higher-quality embeddings at the index step improve end-to-end answer quality.
- Accessible through AresGen so you can build embedding and retrieval workflows in the same workspace where you develop, test, and deploy AI features — without a separate Voyage AI account.
Available in
Use voyage-3 inside these AresGen tools
AI Chat
Use voyage-3 embeddings to ground AI chat answers with retrieved content from your own documents and knowledge bases.
ExploreMarketing Automation
Power semantic retrieval across campaign assets, product content, and customer data with voyage-3 embeddings.
ExploreAI Writer
Retrieve semantically relevant content from your knowledge base to inform and ground AI-generated writing.
ExploreWhen to pick voyage-3 over OpenAI text-embedding-3-large
voyage-3 is a retrieval-focused embedding model from Voyage AI with a 32K token context window, suited to teams that need high retrieval accuracy in semantic search and RAG pipelines and want to embed longer document passages without extensive chunking. OpenAI text-embedding-3-large is a general-purpose embedding model from OpenAI for teams already within the OpenAI ecosystem. Choose voyage-3 when retrieval accuracy and long-context embedding are the primary requirements; browse the full models catalog to compare other available embedding options.
- All models
Browse the full model catalog to compare other embedding models available in AresGen.
Frequently asked
What does voyage-3 produce?
Does voyage-3 support function calling?
What is the context window for voyage-3?
Who makes voyage-3?
Related models
Explore related models
Claude Opus 4.7
Anthropic's most capable model — pair retrieval via voyage-3 embeddings with high-quality generation.
Learn moreGPT-5
OpenAI's flagship model — use with voyage-3 embeddings in RAG pipelines for grounded, accurate responses.
Learn moreGemini Pro 2.5
Google's 1M-context model — combine with voyage-3 embeddings for retrieval-augmented workflows at scale.
Learn moreGet started today.
Free for 7 days. No credit card. Bring your team — or just your first prompt.