We’re excited to announce the public preview of DiskANN with filtering in vCore-based Azure Cosmos DB for MongoDB. With DiskANN, you can perform large-scale vector searches while reducing memory usage by up to 10 times compared to other algorithms like IVFFlat and HNSW—making recommendation engines, fraud detection, and other AI applications more cost-effective and scalable. Enjoy low-latency, high-availability performance, even with billions of vectors.