Voyage AI

What Voyage AI does

Voyage AI develops embedding models and rerankers used to power retrieval-augmented generation (RAG) and semantic search applications. The company was founded in 2023 by Stanford professor Tengyu Ma and is headquartered in Palo Alto, California. Its product portfolio includes the voyage-4 series of text embedding models using Mixture of Experts architecture, domain-specific models for finance (voyage-finance-2), legal (voyage-law-2), and code (voyage-code-3), the voyage-multimodal-3.5 model supporting text, image, and video, and the rerank-2.5 family of cross-encoder rerankers with instruction-following capabilities. Technical differentiators include 3x-8x shorter vectors than competitors, a 32K-token commercial context length, and top ranking on public retrieval benchmarks including MTEB and RTEB.

The company operates an API-first, product-led go-to-market motion with a 200M-token free tier and usage-based pricing thereafter (e.g., $0.12 per million tokens for voyage-4-large). Customers range from individual developers integrating via Python/TypeScript SDKs to enterprises procuring through AWS Marketplace and Azure Marketplace for in-VPC deployments, and platform integrations that embed Voyage models inside MongoDB Atlas Vector Search and Snowflake Cortex AI. Named enterprise customers include Harvey, Replit, SK Telecom, Vanta, and Snowflake, with the broader MongoDB Atlas customer base of 62,500+ providing distribution leverage.

Voyage AI raised $28 million in total funding, including a $20 million Series A led by Snowflake in October 2024, before being acquired by MongoDB in February 2025 for approximately $220 million. At the time of acquisition the company had approximately 19 employees, implying a valuation of roughly $11.6 million per employee. The company is now a wholly-owned subsidiary of MongoDB, with its models forming a core component of MongoDB's GenAI strategy and Atlas Vector Search product surface.

Name

Voyage AI

Legal name

Voyage AI Innovations, Inc.

Website

https://voyageai.com

Company type

Private

Founded year

2023

Operating status

Operating

Headcount range

11–50 employees

Short description

Voyage AI develops proprietary embedding models and rerankers that power retrieval-augmented generation and semantic search. Founded in 2023 by Stanford professor Tengyu Ma, it serves AI developers and enterprises via API, cloud marketplaces, and integrations with MongoDB Atlas and Snowflake Cortex AI; acquired by MongoDB in February 2025 for ~$220 million.

Ownership category

akta.pro rank

Product category

AI Embedding Models and Rerankers

NAICS

Web Search Portals and All Other Information Services (519290)

SIC

Services-Computer Integrated Systems Design (7373)

akta.pro primary industry

Generative AI & LLM Solutions Services (RAG, Agents, Copilots) (BPAEAHAG)

Keywords

• Embedding models
• Semantic search
• Retrieval augmented generation
• Text embeddings
• Reranking models

Headquarters

HQ city

Palo Alto

HQ country

United States

HQ region

North America

Offices1 record

Markets served

GTM type

B2B

Offering type

Software

Cost components

Technology or R&D, Personnel, Infrastructure, Marketing or Sales, Operations

Revenue model

1. API Usage-Based Pricing: Voyage AI charges customers based on the number of tokens processed through their API. Each model has a free tier allocation (200M tokens for most models), after which usage is billed per token. This consumption-based model aligns cost with value delivered and scales with customer usage.
2. Cloud Marketplace (AWS/Azure): Customers can deploy Voyage AI models through AWS Marketplace and Azure Marketplace, with hourly billing that includes both software pricing and infrastructure costs. This provides an alternative revenue channel for enterprise customers preferring marketplace procurement.

Pricing tiers

Go-to-market motion3 records

Distribution channels5 records

Marketing channels6 records

Core offering

Voyage AI develops and provides text and multimodal embedding models and cross-encoder rerankers used to power retrieval-augmented generation (RAG) and semantic search applications. Its models are delivered through a public REST API with Python and TypeScript SDKs, and also deployed via AWS Marketplace, Azure Marketplace, MongoDB Atlas Vector Search, and Snowflake Cortex AI for enterprise customers.

Product overview

Voyage AI is a provider of best-in-class embedding models and rerankers for retrieval-augmented generation (RAG) and semantic search applications. The product portfolio consists of text embedding models (voyage-4 series, voyage-3 series, domain-specific models for finance/legal/code/multilingual), multimodal embedding models (voyage-multimodal-3.5 supporting text/images/video), and reranking models (rerank-2.5 series). These are accessible via a public REST API with Python and TypeScript SDKs, native integrations with popular vector databases and ML frameworks (LangChain, LlamaIndex, Weaviate), cloud marketplace deployments (AWS SageMaker, Azure), and Snowflake Cortex AI integration. The company was acquired by MongoDB in February 2025 for approximately $220 million.

Differentiator

Problem solved

Functional benefit

Products and services

• Embedding Models (Text)
• Rerankers
• voyage-multimodal-3.5
• Voyage AI REST API
• Batch API

Quantifiable outcome

• 45% higher read throughput vs MongoDB 8.0 for AI workloads
• +3 more outcomes

Named customers5 records

Segments5 records

Ideal customer profiles3 records

Technology focussed Yes

API detail

Has API

Yes

API docs

API detail

Core technology

AI maturity

App detail

Integration18 records

AI capability6 records

Feature7 records

Partnerships

Three partnerships are on record, tiered flagship and core.

• MongoDBflagshipOthers · 1 February 2025MongoDB acquired Voyage AI in February 2025 for approximately $220 million. The acquisition integrated Voyage AI's embedding models and reranking technology directly into MongoDB Atlas Vector Search, creating a unified AI database platform. Voyage AI became a core component of MongoDB's GenAI strategy.
• LangChaincoreTechnology or IntegrationVoyage AI embeddings and rerankers are integrated with LangChain as official embedding and reranking providers, enabling developers to easily use Voyage models in LangChain-based RAG applications.
• LlamaIndexcoreTechnology or IntegrationVoyage AI provides official embedding and reranking integrations with LlamaIndex, allowing developers to use Voyage models as node postprocessors and embedding functions within LlamaIndex workflows.

Scale indicators5 records

Recent moves6 records

Expansion highlights5 records

Others

• Weaviate: Weaviate is an open-source vector database with a native Voyage vectorizer and reranker module, serving as both a distribution partner and an adjacent infrastructure player in the RAG stack.
• Pinecone: Pinecone is a leading managed vector database that integrates Voyage as one of several embedding providers; it is an ecosystem participant rather than a direct embedding competitor, but its roadmap choices shape Voyage's distribution.

Direct peers

• Cohere: Cohere provides Embed v3 models and a dedicated Rerank API that are the closest direct substitutes to Voyage's text embeddings and rerankers across enterprise RAG deployments.
• Nomic AI: Nomic AI builds the Atlas embedding model family and Embed API, positioning it as a direct open-weight and hosted alternative to Voyage's general-purpose and long-context embedding offerings.
• OpenAI: OpenAI offers text-embedding-3-small and text-embedding-3-large models via API, directly competing with Voyage's embedding models on retrieval quality and price for RAG and semantic search workloads.
• Jina AI: Jina AI offers embedding, reranker, and reader models targeted at enterprise search and RAG, with overlapping developer-first GTM and comparable product scope to Voyage.

Emerging players

• Together AI: Together AI provides hosted inference for open-source embedding models (including BGE-family) and is a meaningful emerging alternative for cost-sensitive RAG workloads overlapping with Voyage's audience.
• BAAI (BGE): BAAI develops the widely adopted BGE open-source embedding family, which represents the primary open-source competitive threat to Voyage in general-purpose and multilingual retrieval.

Broad incumbents

• Mistral AI: Mistral offers embedding models alongside its flagship LLMs, presenting a bundled alternative for enterprises that prefer to consolidate retrieval and generation providers.
• Google (Vertex AI Embeddings): Google offers text-embedding and multimodal embedding models through Vertex AI as part of a broader cloud AI portfolio, competing with Voyage on retrieval quality while bundling inside a full enterprise AI stack.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat5 records

Key risks6 records

Key highlights7 records

Customer concentration

• Xx.com/VoyageAI
• GitHubgithub.com/voyage-ai
• LinkedInlinkedin.com/company/voyageai

Compliance5 records

Revenue estimate

Valuation estimate

Number of profiles

Profiles1 record

Funding overview

Funding rounds1 record

Investors11 records

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Voyage AI