Database competitor Couchbase released benchmark test results on October 23, 2025, claiming its Hyperscale Vector Index (HVI) is over 350 times faster than MongoDB Atlas for billion-vector scale workloads. The tests, conducted using the VectorDBBench tool, asserted that Couchbase achieved over 700 queries per second with high accuracy, a significant performance advantage in a key technology for building AI applications. The announcement represents a direct competitive challenge to MongoDB in the crucial and rapidly growing market for AI and generative AI application infrastructure. Couchbase's CEO, BJ Schaknowski, stated that such performance differences are critical for enterprises making long-term infrastructure decisions for their AI initiatives. No market reaction or response from MongoDB has been noted in the initial reports.