[pgvector](https://github.com/pgvector/pgvector/), an open-source PostgreSQL extension that provides vector similarity search capabilities, has released [v0.7.0](https://github.com/pgvector/pgvector/releases/tag/v0.7.0). This new release includes many new functional and performance features for supporting vector similarity search workloads in PostgreSQL.
This latest version of pgvector adds new vector types, including `halfvec` (2-byte floats; indexing up to 4,000 dimensions) and `sparsevec` (indexing up to 1,000 nonzero dimensions), and includes indexing support for binary vectors using the `bit` type (indexing up to 64,000 dimensions). Additionally, this release adds support for quantizing vectors using expression indexes, including from 4-byte to 2-byte floats and binary quantization using `binary_quantize` function. pgvector 0.7.0 also adds new distance functions, including `hamming_distance` and `jaccard_distance` for `bit` vectors, and now supports HNSW indexing for L1 distance operations. pgvector 0.7.0 also includes additional support for SIMD with CPU dispatching for Linux x86-64 architectures. For more information, please see the [CHANGELOG for 0.7.0](https://github.com/pgvector/pgvector/blob/master/CHANGELOG.md#070-2024-04-29): [https://github.com/pgvector/pgvector/blob/master/CHANGELOG.md#070-2024-04-29](https://github.com/pgvector/pgvector/blob/master/CHANGELOG.md#070-2024-04-29) For more information about pgvector, including how to get started, please visit the [project repository on GitHub](https://github.com/pgvector/pgvector): [https://github.com/pgvector/pgvector](https://github.com/pgvector/pgvector)
