Weaviate 1.37 Release

2026-04-23 · Source: Weaviate Blog · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Data Science & Analytics, Software Development & Engineering · Depth: Intermediate, long

Summary

Weaviate "v1.37", released on April 23, 2026, introduces significant enhancements for AI agent integration, text analysis, and large-scale data management. Key preview features include a built-in Model Context Protocol (MCP) server, enabling LLMs and IDEs like Claude Code and Cursor to natively interact with the database for agentic workflows. Extensible Tokenizers offer accent folding and custom stopword presets, alongside a new tokenize endpoint for transparency. Diversity Search (MMR) provides query-time reranking to reduce redundant vector search results, while Query Profiling offers per-shard timing breakdowns for performance diagnostics. Additionally, the release brings Incremental Backups for efficient management of massive collections, Gemini audio support to the "multi2vec-google" module for multimodal embeddings, and a BlobHash property type that stores only a SHA-256 hash of media content, reducing storage overhead. Numerous performance improvements and bug fixes are also included.

Key takeaway

For AI Engineers building agentic workflows, Weaviate "v1.37" significantly streamlines database interaction. You should enable the new MCP Server to allow LLMs and IDEs to natively inspect schemas, run hybrid searches, and write data, transforming Weaviate into active long-term memory. Additionally, consider implementing incremental backups for massive collections and MMR diversity search to improve result quality, ensuring your agent systems are both efficient and effective.

Key insights

Weaviate "v1.37" transforms the vector database into an active AI agent memory and enhances core search, backup, and multimodal capabilities.

Principles

• Direct database interaction empowers AI agents for complex workflows.
• Diversity reranking enhances vector search result utility.

Method

Enable the MCP server with "MCP_SERVER_ENABLED: 'true'" to expose database tools to AI agents. Create incremental backups by chaining "client.backup.create" calls with "incremental_base_backup_id".

In practice

• Integrate Weaviate as long-term memory for LLM agentic workflows.
• Implement MMR diversity search to prevent redundant vector results.

Topics

• Weaviate
• AI Agents
• Model Context Protocol
• Vector Search
• Incremental Backups
• Multimodal Embeddings

Best for: AI Architect, NLP Engineer, Computer Vision Engineer, AI Engineer, Machine Learning Engineer, MLOps Engineer

Related on AIssential

• Build a Coding Assistant with Weaviate MCP: RAG over Code & Docs — Weaviate's built-in MCP server simplifies RAG for coding assistants, enabling hybrid search over code and docs.
• Presentation: AI Agents to Make Sense of Data at OpenAI — OpenAI's Kepler agent uses advanced AI techniques to automate complex data analysis across vast datasets, improving data accessibility and accuracy.
• Engram: Memory by Weaviate — Agentic applications require actively maintained, structured memory for effective learning and personalization.
• Hermes Agent v0.15! Huge New Updates: Agent Swarms, Tool Search, NEW Models, & More! — Hermes Agent's velocity update enhances autonomous AI workflows through lazy-loaded tools, agent swarms, and expanded model support, boosting efficiency and scalability.
• Before You Build an AI Agent, Read This — Effective AI agent development prioritizes foundational workflows and adaptive retrieval over static, single-tool approaches.
• memweave: Zero-Infra AI Agent Memory with Markdown and SQLite — No Vector Database Required — LLM agent memory can be managed effectively using version-controlled Markdown files and a local SQLite index.

See Counsel's argued verdicts on the open AI decisions leaders are weighing →

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Weaviate Blog.