Nexus is GA. Enterprise knowledge, compiled once, query-ready for agents. On Sierra's τ-Knowledge: GPT-5.2 +12% accuracy, -80% cost. GPT-5.5 same accuracy, -77% cost.
Read the full announcement: pinecone.io/blog/pinecone-…
Why does filtered vector search break HNSW? And how do you keep an index small enough to share memory with Postgres?
Jörg Schad takes us under the hood of pg_pinecone, our open-source Postgres extension. Watch now: youtu.be/uU6TEeaIh3o
Meet pg_pinecone: a new open-source Postgres extension built on the search engine Pinecone runs at scale. A smaller index, faster builds, filters during search. You get the same SQL you're used to and search runs inside your Postgres database.
Install pg_pinecone:
Chroma Cloud is winding down on December 22.
If your app runs on Chroma Cloud and you need a new managed vector database, our new guide walks through migrating your Chroma collection to Pinecone Database.
Read the step by step instructions:
Hybrid search used to take three indexes. Now it takes one.
New in the Pinecone changelog:
→ Hybrid search in one index
→ BYOC is GA, now with on-demand
→ VQ-Bench for vector quantizers
→ Cost and usage export in FOCUS format
Watch the full changelog 👇