So, I’m excited to share that we will be introducing full-text search and vector search in MongoDB Community Edition later this year, making it even easier for developers to quickly experiment with new features and streamlining end-to-end software development workflows when building AI applications. These new capabilities also enable support for customers who want to run AI-powered apps on devices or on-premises.
While I am unable to give any specific dates as these things are subject to change, I am pleased to say that we will be kicking off the early access program for MongoDB Search on Community most likely next month. If you would be interested in participating and offering feedback please sign up for the program here.
The early access program will run for a few months as a Private Preview while we work on eliminating technical limitations and incorporating feedback into a public preview release. The Private Preview will be similar in scope to the current local development experience with the Atlas CLI in that we will not support sharded clusters, only a single search node is supported, and a few other limits. The $search, $searchMeta, and $vectorSearch aggregation stages should be functionally identical to what is in Atlas today.
Public Preview provides index lifecycle management and querying functionality for both unsharded and sharded collections on a minimal set of supported platforms. The Public Preview is not recommended for production usage.
GA extends Public Preview capabilities to offer better scalability, improve query availability and performance, and open source mongot (the search process) under the SSPL. We will make the decision to cut a GA release when our quality, security, feature-level requirements are satisfied. This is a top priority for us and we are working diligently to bring search and vector search to MongoDB Community as soon as possible.