voyage-4-large
Voyage’s highest-quality general and multilingual retrieval embedding.
voyage-4-large is a general-purpose, multilingual embedding model optimised for retrieval quality. Matryoshka learning and quantization-aware training let it emit 2048, 1024, 512 or 256-dimension vectors with multiple quantization options.
- Provider
- Voyage AI (MongoDB)
- Type
- Embedding model
- Released
- Jul 27, 2026
- Context window
- 32K tokens
- Price
- $0.12 per 1M input tokens
- Input
- text
- Output
- embedding
- Open weights
- No
Best for
- Research
- Long context
- Enterprise
Strengths
- Matryoshka dimensions (256–2048)
- Quantized embeddings
- 32K-token inputs
More from Voyage AI (MongoDB)
- voyage-code-4 — Voyage AI (MongoDB). Code-retrieval embedding built for coding agents.
- voyage-multimodal-3.5 — Voyage AI (MongoDB). Embeds text, images and interleaved content into one space.
Tools that use it
- MongoDB Atlas Vector Search — MongoDB. Vector search inside your operational MongoDB database.