MongoDB, Inc. (NASDAQ: MDB) announced several new database capabilities at its MongoDB.local Bengaluru event on June 30, 2026, aimed at improving AI retrieval accuracy and expanding access to on-premises and private cloud environments.

The company launched three retrieval-focused features. Native Reranking in MongoDB Atlas, now in public preview, runs inside the database's aggregation pipeline and is said to improve retrieval quality by up to 30%, based on Voyage instruction-following rerankers on the MAIR benchmark. Voyage Context 4, now generally available, is an embedding model designed to process long documents in full context. Hybrid Search, also now generally available, combines full-text and vector search in a single database query.

MongoDB also made Search and Vector Search generally available as an add-on for MongoDB Enterprise Advanced, bringing those capabilities to on-premises, private cloud, and hybrid deployments. The company said more than 20 large banks and financial institutions had been evaluating the feature ahead of its release. The same capabilities are now generally available for MongoDB Community Edition at no cost.

"Whether you're running in the cloud, private cloud, or behind a firewall, MongoDB gives you the same production-grade retrieval capabilities wherever your data lives," said Ben Cefalo, Chief Product Officer, Core Products, at MongoDB.

Additional releases included Apache Iceberg support for MongoDB Atlas Stream Processing, next-generation infrastructure for Atlas M30+ dedicated clusters on AWS, and asymmetric Search Node deployment for multi-region clusters, which the company said could lower search node costs by 25–40%.

Separately, MongoDB announced a plan to train two million builders in India by 2030 through its MongoDB for Academia program, in partnership with the All India Council for Technical Education, HCL GUVI, and the ICT Academy of Kerala. The program has reached more than 650,000 students since 2023. The company also launched a startup competition offering $50,000 in MongoDB Atlas credits along with travel and go-to-market opportunities.