Weaviate 1.37 ships built-in MCP server, turning the vector DB into agent memory
CShorten30 · x · 2026-10-20
Weaviate v1.37 is out, focused on plugging the vector database directly into agent workflows.
Four preview features:
- Built-in MCP Server: MCP-aware clients (Claude Code, Cursor, VS Code) can natively inspect schemas, run hybrid searches, and write data back — no glue code, turning Weaviate into active long-term memory for agentic workflows
- Extensible tokenizers with accent folding and custom stopword presets
- Diversity Search (MMR) to cut redundancy in vector results
- Query Profiling with per-shard timing breakdowns
Also included: incremental backups for massive collections, Gemini audio in the multi2vec-google module, and a new BlobHash property type that stores only a SHA-256 hash instead of raw media — assets stay in object storage and re-vectorization happens only when content actually changes.
More from coding & agent
- OpenAI's Decisions API hits GA: near real-time routing up to 10x faster than GPT-6 Luna — OpenAIDevs · 2026-10-23
- Ruff author warns: upgrade Node/Bun now — --use-system-ca still accepts the eMudhra cert — charliermarsh · 2026-10-23
- Proof of Agent: paste one prompt and your AI mines, grades, and self-governs — adamamcbride · 2026-10-23
- DeepLearning.AI weekly: GPT-6.1 hits 97.9% byte parity on COBOL migration — DeepLearningAI · 2026-10-23
- New data marketplace lets vendors list repos; coding agents access via MCP server — testingcatalog · 2026-10-23
- Tie code approvals to content hashes so agent 'one more small fix' voids review — Ambitious_Answer6857 · 2026-10-23