Voyage AI
AI NativeVoyage is a team of leading AI researchers and engineers, building embedding models for better retrieval and RAG.
MarketsFoundation Models
Published Oct 15, 2026
No authoritative market share figures were found. Signals of adoption include the vendor's acquisition by MongoDB, which provides a strategic distribution channel, and availability via marketplace deployment channels.
| Starting Price | Usage-based per-token pricing with multiple model tiers; vendor positions lower-cost lower-tier options for cost-sensitive workloads | Per-token pricing with broad ecosystem adoption and straightforward billing | Per-character / per-image / per-second pricing models for multimodal/video embeddings; complexity varies by SKU |
| Most Appropriate For | RAG/search teams optimizing retrieval quality and vector DB cost; long-context and multimodal needs | Teams standardized on OpenAI for ecosystem and tooling integration | Organizations already deep in Google Cloud; multimodal and video embedding needs |
| Key Strength | Matryoshka + quantization options; emphasis on reranker capability and long-context support | Simple integration; widespread usage; clear documentation and ecosystem tools | Explicit multimodal and video embedding SKUs; tight GCP integration |
| Key Weakness | Limited third-party review signal and less publicly documented enterprise trust materials | Generally positioned at higher per-token price tiers for some models; embeddings-only focus for many SKUs | Pricing complexity and cloud lock-in considerations |
Voyage vs OpenAI: Voyage publishes competitive per-token pricing for several model tiers and positions itself as cost-competitive on lower-cost model options. However, OpenAI remains attractive due to simplicity, broad adoption, and ecosystem integration.
Voyage vs Cohere: Cohere competes in enterprise positioning and offers multiple model families; Voyage emphasizes retrieval-specific benchmarks and reranker performance.
Voyage vs Google Vertex AI (multimodal/video): Voyage targets text+image+video-frame retrieval use cases. Google provides explicit multimodal/video embedding SKUs and may suit teams already deploying on GCP.