MongoDB is more tightly integrating the embedding and reranking model technology it acquired through last year’s acquisition of Voyage AI with its database development platform.
MongoDB is extending the AI capabilities of its database and application development platform with newly integrated embedded and reranking models, which the company says will improve the accuracy of AI applications as they move from development to production.
The company says this new feature is based on the technology MongoDB acquired when it acquired Voyage AI in February 2025, and provides a unified data intelligence layer for production AI, enabling developers to build and operate advanced AI applications at scale with minimal risk of illusion and without the need to move or replicate data.
Today, with the MongoDB.local San Francisco event, MongoDB also announced an expansion of MongoDB for Startups, a program that provides startups with technical expertise, financial credit, and other resources to build AI software using a technology stack powered by MongoDB Atlas.
[Related: MongoDB Names Former Cloudflare Exec To Take Over As CEO]
“Over the past few months, we’ve spent time with countless customers, founders, executives at large companies, platform teams, and developers, not to make recommendations, but to understand what goes wrong as AI moves from prototype to production,” Ben Cefalo, MongoDB’s senior vice president and head of core products and Atlas Foundational Services, said at a press conference ahead of today’s MongoDB.local San Francisco event.
“Those conversations rarely start with an AI model. They start with very practical questions like, ‘How do I prepare my data? How do I maintain performance as I scale? How do I ensure accuracy of my results? How do I avoid stringing together five different systems and extensions just to ship something? What’s the ROI?'” Cefalo says.
While MongoDB is generally considered one of the leading next-generation database systems, the company has recently positioned its software, specifically its MongoDB Atlas cloud-native database, as the foundation of the technology stack needed to develop and run AI applications.
In 2024, the company launched the MongoDB AI Application Program (MAAP). Through this program, the company partners with cloud hyperscalers, large-scale language model developers, AI development tool providers, system integrators and consulting partners to provide technology stacks and reference architectures for building AI systems.
Cefalo pointed to features within the MongoDB platform that support production AI systems at scale, including support for structured, semi-structured, and unstructured data. Real-time operational data. Data accuracy control. Vector search and hybrid search capabilities. and automatic embedding.
“At the end of the day, this job all comes down to one thing: helping builders build,” Cefalo said. “Our database, platform, and industry-leading AI capabilities all help us turn ideas into working systems.”
According to the company, Voyage AI’s embedding and re-ranking models integrate with the MongoDB core database to provide an integrated data intelligence layer to production AI. By adding these models to the MongoDB platform infrastructure, developers can reduce the risk of illusion and build and operate complex AI applications at scale without the need to move or replicate data.
new nautical model
MongoDB announced the general availability of the new Voyage 4 model series, including the general-purpose Voyage-4 embedded model, the flagship Voyage-4 large-scale language model for improved search accuracy, Voyage-4-lite, and the openweight Voyage-4-nano for local development and testing and on-device applications.
Additionally, a new Voyage-Multimodal-3.5 model with expanded support for interleaved text and images is now generally available.
The company also announced a public preview of automatic embedding of MongoDB Vector Search in MongoDB Community edition and will be available soon in MongoDB Atlas. Additionally, the Atlas Embedding and Reranking API, which exposes Voyage AI models natively within Atlas, is now generally available.
“This announcement addresses one of the most difficult parts of building an AI application: managing the search and embedding models and pushing them to the database,” Frank Liu, Voyage AI product manager, said in a press conference.
“For builders, this means they can manage all their data, embedding, and retrieval in one place. There is less friction when retrieving data. [an] “This is all designed to combine with the rest of MongoDB’s features, such as automatic embedding, vector search, and a more integrated developer experience. So when you choose MongoDB for your AI workloads, you’re not just getting a database, you’re getting a search stack that can sustain your ambitions as a developer,” said Liu.
Extending MongoDB for startups
According to MongoDB, companies participating in the MongoDB for Startups program now have a combined valuation of more than $200 billion. The overall goal of this program is to provide startups with a complete infrastructure stack, allowing them to avoid having to spend time making infrastructure decisions.
Under this expansion, the program’s ecosystem of supporting IT vendors, including AI workflow platform developer Temporal and generative AI platform provider Fireworks AI, will provide startups with match credits to realize benefits across content, collaborative events, and other complementary technologies.
“Our goal is to transform MongoDB for Startups from a one-way benefits marketplace to a two-way ecosystem where both partners and startups benefit from participating in the program,” Suraj Patel, vice president of MongoDB Ventures & Corporate Development, said in a press conference.
“This mutual partnership with MongoDB allows us to reach a community of developers who value a strong data foundation,” Temporal CEO Samar Abbas said in a statement. “We look forward to building a collaborative ecosystem that simplifies complexity for founders pushing the boundaries of distributed systems and workflow orchestration.”
“By joining this program, we guarantee: [startup] Founders who choose MongoDB will have easy access to our high-performance inference engine, creating a seamless path to scale their AI ambitions together,” Lin Qiao, co-founder and CEO of Fireworks, also said in a statement.
MongoDB is also putting more emphasis on recruiting more startups from San Francisco and the Bay Area, with plans to host more than 50 local events with a specific focus on AI startups next year.
