🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

✨A zero to hero guide to create, store and query OpenAI embeddings with PostgreSQL and pgvector. Learn how to build a chatbot to answer questions about blog posts using PostgreSQL as a vector database for OpenAI embeddings. timescale.com/blog/postgresq…

Jonathan Katz (Jonathan S. Katz): Vectors are the new JSON in PostgreSQL postgr.es/p/5N4

pgvector is now available for Google Cloud SQL and AlloyDB 🎉 cloud.google.com/blog/products/…

🧠 A developer's guide to the ivfflat index in pgvector Learn how it enables approximate nearest neighbor search when using PostgreSQL as a vector database, plus best practices (and gotchas!) when using it for similarity search over OpenAI embeddings. timescale.com/blog/nearest-n…

pgvector 0.5.1 is out with much faster HNSW build times thanks to Heikki Linnakangas 🔥

New episode: "pgvector" Michael is joined by Jonathan Katz (Jonathan S. Katz), PostgreSQL Core Team member, Product Manager Amazon Web Services, and pgvector contributor. They cover what it is, some tuning guidance, and a look ahead. 🎧 postgres.fm/episodes/pgvec… 📺 youtube.com/watch?v=vvImP6…

pgvector 0.6.0 is out 🎉

Postgres community delivers again 💚 pgvector 0.6.0: 30x faster with parallel index builds Link: supabase.com/blog/pgvector-…

pgvector 0.6.2 is now available, improving performance for HNSW index builds with a large number of parallel workers 🔥

Support for indexing binary vectors has been merged and will be available in 0.7.0 🔥

Support for float16 vectors has been merged 🔥 Store and/or index your vectors at half-precision to save space and memory

Jonathan Katz (Jonathan S. Katz): Scalar and binary quantization for pgvector vector search and storage postgr.es/p/6sO

Hybrid search combines full text and semantic search for more relevant results Learn how to implement it using pgvector and OpenAI 🤖 supabase.com/docs/guides/ai…

pgvector now uses intrinsics + CPU dispatching for key float16 operations, providing much better halfvec performance on x86-64 than before 🔥

pgvector 0.7.0 will support indexing vectors at half precision with both HNSW and IVFFlat 🔥

pgvector 0.7.1 is now available with improved performance for on-disk HNSW index builds (thanks to findings from Tomas Vondra) 🔥

pgvector 0.8.0 is now available with improved filtering and performance 🔥