Mongodb from AI4 has announced a wide range of product innovations and an ecosystem expansion for AI partners. By providing an industry-leading embedded model, a fully integrated AI-enabled data platform, and assemble a world-class ecosystem of AI partners, MongoDB offers organizations everywhere the tools to deliver reliable performance, cost-effective AI.
Organizations recognize the business potential of AI. However, according to the 2025 Gartner Gartner Generation and Agent AI Enterprise Application Survey, 68% of IT leaders feel they are struggling to keep up with the rapid pace at which Gen AI tools are deployed, while 37% agreed to their application vendors drive their enterprise application Gen AI strategy. Too many organizations are stuck in the messy middle with AI implementations and see some benefits, but it's not enough to guarantee wider adoption.
Companies have expressed that this gap in AI adoption (a barrier for both developers and enterprises) is due to the complexity of the AI stack, the importance and challenges of achieving mission-critical applications. Price performance concerns It appears on a large scale. To address these issues, MongoDB has invested in streamlining the AI stack and continues to introduce more performant, more cost-effective models. Customers can integrate the latest embedded models of Voyage AI with their MongoDB database infrastructure. MongoDB has increased interoperability with industry-leading AI frameworks by launching the MongoDB MCP server to allow agents to access tools and data, and expanding its comprehensive AI partner ecosystem to allow more choice of developers.
advertisement
These features are gaining great momentum among developers building next-generation AI applications. Enterprise AI adopters such as Vonage, LGU+ and Financial Times employ roughly 8,000 startups, including using timekeeping startups Laurel and Mercor, to use AI to match talent with opportunities. Meanwhile, over 200,000 new developers register with Mongodb Atlas every month.
“Databases are more central than ever for the technology stack in the age of AI. Modern AI applications require databases that combine integrated vector search with advanced features such as best-in-class AI models. “These systems also require scalability, security and flexibility to support production applications as production applications evolve and usage increases by integrating AI data stacks and building cutting-edge AI ecosystems.”
Accelerate AI innovation with enhanced product features
Voyage AI by Mongodb recently introduced an industry-leading embedded model designed to unlock AI accuracy at a lower cost.
- Context recognition embedding for better search: The new Voyage-Context-3 model brings a breakthrough in AI accuracy and efficiency. Metadata hacks, LLM summary, or required pipeline exercises capture the full document context (metadata hacks, LLM summary, or pipeline exercises) that lead to more relevant results and reduce sensitivity to chunk size. It acts as a standard embedding drop-in exchange for RAG applications.
- New highs in model performance: The latest general purpose models, Voyage-3.5 and Voyage-3.5-Lite, raise the bar with search quality, providing industry-leading accuracy and price performance.
- Re-ranking following guidance to improve accuracy: With Rerank-2.5 and Rerank-2.5-Lite, developers can now use instructions to guide the Reranking process to unlock greater search accuracy. These models outperform their competitors in a comprehensive set of benchmarks.
MongoDB recently introduced a MongoDB Model Context Protocol (MCP) server in Public Preview. This server standardizes MongoDB deployments to connect directly to popular tools such as Visual Studio Code, Anthropic's Claude, Cursor, and Windsurf's Github Copilot. Developers interact with data using natural language, manage database operations, and streamline application development powered by Mongodb's AI.
Since its launch in public preview, MongoDB MCP servers have grown rapidly in popularity, with thousands of users being built on Mongodb each week. Mongodb is also very interested in large enterprise customers considering incorporating MCP as part of it Agent application stack.
“Many organizations struggle to scale up AI because the models themselves cannot withstand the tasks. They lack the accuracy needed to please their customers, and are complex to fine-tuning and integration, and become too expensive at scale. “The quality of embedded and reranking models is often the difference between promising prototypes and AI applications that produce meaningful results. That's why we focus on building better performance, cost and easy-to-use models so that developers can bring and adopt AI applications into the real world.”
“As more companies deploy and expand their AI applications and agents, the demand for accurate power and reduced latency continues to increase,” said Jason Andersen, vice president and principal analyst at Moor Insights and Strategy. “By thinking and integrating an AI data stack with advanced vector search and embedding integrated into the core database platform, MongoDB is taking on these challenges while reducing developer complexity.”
Expanding the MongoDB AI ecosystem
MongoDB also expanded its ecosystem of AI partners to help customers build and deploy AI applications faster.
- Enhanced evaluation features: Galileo, a leading AI reliability and observability platform, is a member of MongoDB partner Ecosystem and is designed to provide flexibility and choice to its customers. With continuous evaluation and monitoring, Galileo enables reliable deployment of AI applications and agents built on MongoDB.
- Resilient and Scalable AI Applications: A leading open source, durable execution platform, Tumeal is also a member of the MongoDB partner ecosystem. In time, developers can coordinate trusted AI use cases built on MongoDB, including agents, RAG, and context engineering pipelines that manage and provide dynamic, structured contexts at runtime. With the durable execution of Tuperal, developers do not need to write plumbing code for resilience or scale. AI applications recover seamlessly across failures, reliably run long periods, easily handle external interactions and scale horizontally. Developers can also visualize every step of their AI workflow and quickly debug live issues. These partner features greatly expand MongoDB's AI ecosystem to develop AI applications.
- Streamlined AI workflow: Mongodb Partnership Running Chain By streamlining development and unlocking the value of real-time, unique data from customers, it redefines the way developers build AI applications and agent-based systems. Recent advances include the introduction of GraphRag using Mongodb Atlas. This increases transparency into the search process, promotes trust, and provides better explanationability for LLMS responses. Another advancement is natural language queries in MongoDB. This allows agent applications to interact directly with MongoDB data. These integrations allow developers to build reliable, sophisticated AI solutions, from highly searched, high-score (RAG) systems to autonomous agents that can query data and perform advanced searches.
“As organizations deploy AI applications and agents into production, accuracy and reliability are extremely important,” says Vikram Chatterji, CEO and co-founder of Galileo. “By formally joining Mongodb's AI ecosystem, Mongodb and Galileo can enable customers to deploy trustworthy AI applications that transform their businesses with less friction.”
“Building production-ready agent AI means ensuring systems can withstand real-world reliability and scale challenges. “Through our partnership with MongoDB, Develolal will confidently coordinate durable, horizontally scalable AI systems, enabling engineering teams to build applications that our customers can expect.” “As AI agents take on increasingly complex tasks, access to diverse and relevant data becomes essential,” says Harrison Chase, CEO and co-founder of Langchain. “Integration with MongoDB, including features such as GraphRag and natural language queries, equips developers with the tools they need to build and deploy complex, future agent AI applications based on relevant, reliable data.”
