Vector Search Limitations

This document describes the known limitations of TiDB vector search.

Vector data type limitations

• Each vector supports up to 16383 dimensions.
• Vector data types cannot store NaN, Infinity, or -Infinity values.
• Vector data types cannot store double-precision floating-point numbers. If you insert or store double-precision floating-point numbers in vector columns, TiDB converts them to single-precision floating-point numbers.
• Vector columns cannot be used as primary keys or as part of a primary key.
• Vector columns cannot be used as unique indexes or as part of a unique index.
• Vector columns cannot be used as partition keys or as part of a partition key.
• Currently, TiDB does not support modifying a vector column to other data types (such as JSON and VARCHAR).

Vector index limitations

See Vector search restrictions.

Compatibility with TiDB tools

• Make sure that you are using v8.4.0 or a later version of BR to back up and restore data. Restoring tables with vector data types to TiDB clusters earlier than v8.4.0 is not supported.
• TiDB Data Migration (DM) does not support migrating or replicating MySQL 9.0 vector data types to TiDB.
• When TiCDC replicates vector data to a downstream that does not support vector data types, it will change the vector data types to another type. For more information, see Compatibility with vector data types.

Feedback

We value your feedback and are always here to help:

• Join our Discord