At MongoDB.local San Francisco on January 15, 2026, MongoDB announced an industry-first unification of its core database with Voyage AI’s embedding and reranking models. The launch introduces five new embedding models—voyage-4, voyage-4-large, voyage-4-lite, voyage-4-nano and voyage-multimodal-3.5—directly embedded into MongoDB’s platform infrastructure. This integration eliminates the need for separate vector stores or external model services, enabling developers to build, deploy and operate AI-powered applications at scale without duplicating data or introducing additional latency. Automated Embedding for MongoDB Community Vector Search and AI-powered assistants for Compass and Atlas Data Explorer will be available in public preview, supporting major drivers and AI frameworks such as JavaScript, Python, LangChain and LangGraph.

The general availability of the Voyage 4 series positions MongoDB at the top of the public RTEB leaderboard, outperforming competitors on retrieval accuracy while reducing inference cost and latency. The flagship voyage-4-large model offers the highest semantic retrieval precision, while voyage-4-lite and voyage-4 strike an optimized balance between cost and performance. For on-device and local development use cases, the open-weights voyage-4-nano model provides low-footprint embedding capabilities. Concurrently, the voyage-multimodal-3.5 model extends unified processing to interleaved text, images and video, further simplifying context extraction from complex documents such as slides, tables and figures with minimal pre-processing overhead.

MongoDB’s unified data intelligence layer addresses the fragmentation challenge faced by enterprises that previously juggled operational databases, vector stores and multiple pipelines. By handling embedding generation natively—triggered automatically on data insertion, update or query—MongoDB reduces synchronization overhead and operational risk. This streamlined architecture offers faster iteration cycles, lower latency and consistent embedding freshness, ensuring AI applications maintain reliable context as datasets evolve. The platform is already trusted by more than 60,000 customers, including over 75% of the Fortune 100, to power mission-critical workloads with integrated operational data, real-time analytics and AI retrieval services.

Startups Tavily and TinyFish have adopted MongoDB’s AI enhancements to accelerate feature development and scale operations. TinyFish reported that Voyage AI’s Python APIs delivered lightweight, high-performance embeddings at scale, while Tavily highlighted the ability to focus resources on core product innovation rather than infrastructure management. MongoDB’s new AI skills certification program launched concurrently aims to upskill development teams and foster best practices for deploying AI in production. Senior Vice President Fred Roma emphasized that these capabilities reduce complexity and clear the path from prototype to production, setting a new standard for AI-driven application development.