Weaviate Enterprise Edition, 1.40 Release, and Zero to RAG in 20 minutes Webinar

Hello Weaviate Community! 🤗

We’re excited to introduce the Weaviate Enterprise Edition! Run Weaviate on your own infrastructure with additional security, resilience, operational features, and enterprise support. Find the details on the new Enterprise Edition page.

Weaviate 1.40 is also out, with GA for dropping named vectors and 4-bit Rotational Quantization, plus multi-vector embeddings on HFresh and a new PathSeer filter strategy.

New to RAG, or want a quick refresher? Join our webinar on November 19: Zero to RAG in 20 minutes. Register here to save your spot.

Latest AI & tech insights

Explore our recent Weaviate content:

Read

• 🚀 Weaviate 1.40 Release: The full changelog behind this release, including a breaking change to vector configuration defaults to check before you upgrade. Read the blog

• 🧭 PathSeer: Adaptive Filtered Vector Search: PathSeer is a third HNSW filter strategy based on a SIGMOD 2026 paper and contributed by the paper's first author. The post explains how it works and how it compares to sweeping and ACORN. Read the blog

• 🔐 How Weaviate responded to the eMudhra certificate compromise: A task force drawn from every engineering team rolled the whole environment to Go 1.27.3 after eight certificates leaked from the eMudhra certificate authority. Enterprise customers were patched within six hours and the full estate in under 17. Read the blog

Watch

• 🎙 ZeroEntropy with Ghita Houir Alami - Weaviate Podcast #150: Ghita, ZeroEntropy co-founder and now leading Model Research at Notion, joins us to discuss calibrated rerankers, multi-stage retrieval, and why specialized models can make agentic search more efficient. Watch the full podcast

• 🎙 Scaling Discovery with Shangyin Tan - Weaviate Podcast #149: Shangyin explores why long-running AI optimizers plateau, how “forgetting” helps them resume progress, and what multi-agent teams can learn by sharing Git diffs instead of full code. Watch the full podcast

• 🎙 Machine Studying with Jacob X. Li - Weaviate Podcast #148: Jacob, a PhD student at MIT, explains how agents can study a corpus before they know the downstream tasks. That corpus could be a codebase, a stack of drug discovery papers, or system logs. He also covers how the paper measures expertise as performance against a search-compute budget. Watch the full podcast

• 🎙 SkyDiscover - Weaviate Podcast #147: Four PhD students from UC Berkeley's Sky Computing Lab discuss agents that synthesize whole systems, such as key-value stores tuned to a single workload. Their agents reached up to 4x better performance than specialized systems from Microsoft. They also explain how SkySynth catches reward hacking with test-driven or formal verification routes. Watch the full podcast

• 🎙 In-Context Retrieval with Siddharth Gollapudi - Weaviate Podcast #146: Siddharth discusses using attention itself as the retriever over a corpus held in context. The episode covers BlockSearch, a 0.6B in-context retriever, and what breaks when you scale toward 10,000 documents and roughly a million tokens. Watch the full podcast

• 🔌 Introducing the Query Agent MCP Server: The Query Agent now runs as a remote MCP server hosted by Weaviate. Any MCP-compatible client, such as Claude Code, Cursor, or Notion AI, can query your Weaviate Cloud cluster without installing a client library. Watch the video

• 📺 Inside Weaviate's MCP Ecosystem | Webinar: Connor Shorten and Ivan Despot give live demos of the Weaviate Database and Query Agent MCP servers. Alexander Ledbetter from Delegance Brokerage also shares how they built and shipped their own MCP server on Weaviate. Watch the webinar

Product highlights

Check out our latest product updates.

🏢 Weaviate Enterprise Edition: Run Weaviate on your own infrastructure with additional security, resilience, operational features, and enterprise support. Learn more

🚀 Weaviate 1.40 Highlights:

• Drop a named vector (GA): Delete an unused vector's index and embeddings to free disk and index memory without re-creating the collection. Cleanup runs in the background, shard by shard.

• 4-bit Rotational Quantization (GA): Store a 1536-dimension vector in 784 bytes, 7.84x smaller than raw float32.

• Multi-vector embeddings on HFresh: The HFresh index now accepts multi-vector embeddings through MUVERA.

• PathSeer filter strategy: A third filtering option for HNSW, alongside ACORN and sweeping.

• Search REST API (Preview): The Search REST API lets you search with plain JSON over HTTP/1.1, with no generated client and no GraphQL query strings.

• Vector configuration defaults (breaking change): If your server sets DEFAULT_VECTORIZER_MODULE, read Vector configuration defaults before you upgrade.

See the full release blog for the details.

Upcoming events

Join us online or in person for expert-led AI workshops and networking events. Connect, learn, and build with the community!

Online

• 📚 Zero to Rag in 20 minutes | Thursday, November 19. Register here

Hungry for more?

Have a question or want to connect? Join our Weaviate Forum to engage in community conversations.

Follow our Blog, LinkedIn, Twitter, and YouTube for more content updates.

See you next month,

Kamil