MaxCompute - MaxCompute multimodal retrieval capability released
Jul 30 2026
MaxComputeBuilding on its existing offline batch processing and data analysis engine, MaxCompute provides end-to-end capabilities through the native VECTOR data type, the AI_EMBEDDING function, the VECTOR INDEX, and the VECTOR SEARCH Function. This enables users to directly use SQL to complete the end-to-end pipeline from multimodal data warehousing, embedding generation and vectorized storage, index construction, to final offline batch similarity retrieval, especially for scenarios involving massive data.
Content
Target customers: All customers. New features /specifications: Building on its existing offline batch processing and data analysis engine, MaxCompute provides end-to-end capabilities through the native VECTOR data type, the AI_EMBEDDING function, the VECTOR INDEX, and the VECTOR SEARCH Function. This enables users to directly use SQL to complete the end-to-end pipeline from multimodal data warehousing, embedding generation and vectorized storage, index construction, to final offline batch similarity retrieval, especially for scenarios involving massive data.