• MATCHES
  • DETAILS
  • LISTS
  • HUMAN

How to perform semantic vector searches with your AI

Connect Weaviate to your AI client to move beyond keyword matching. Your agent will find information based on the actual intent behind your questions.

Ask AI about this page

Short answer

How can I search for meaning instead of keywords?

Your AI uses vector similarity to find objects in Weaviate that are conceptually related to your query. This allows you to find answers even when the exact words don't match. You get precise context for every search.

Search outcomes

Where your data lands

What happens when your AI queries the vector database.

MATCHES

Semantic hits

The AI identifies objects that share the same mathematical space as your query. You get results that actually make sense in context.

DETAILS

Deep object data

Once a match is found, the AI pulls the specific properties of that object. You get the full context needed to answer your question.

LISTS

Object discovery

The AI can see what exists within your collections. This helps it understand the scope of your data before searching.

HUMAN

Manual review

You decide which semantic matches are actually useful. The AI provides the candidates, but you make the final call on relevance.

The workflow

What your AI does when <the thing arrives>.

The AI follows a specific path to turn your natural language into a mathematical search.

  1. Scan the collection

    The AI first looks at what kind of data is available in your Weaviate instance to understand the schema.

    list_objects

  2. Find similar data

    The agent converts your question into a vector and finds the closest matches in the database.

    search_near_vector

  3. Pull specific info

    After finding a match, the AI grabs the full details of that specific object to provide a complete answer.

    get_object_details

  4. Verify the results

    The AI compares the retrieved data against your original intent to ensure the context is correct.

    get_object_details

Try it

Copy these to start.

Use these prompts to begin interacting with your Weaviate data.

Starting points

These are starting points. Swap out the specific collection names or object types for your own data.

Weaviate Connector

You're all set. Choose your MCP client and follow the setup instructions.

Connector linkhttps://edge.vinkius.com/vk_preview_ewBpKaEj3GWOCB4HdPc1Tag8b8QBrbGJDpCpFWfW/mcp

Claude Desktop

Follow the steps below to connect in seconds.

  1. 1In Claude Desktop, open Settings → Connectors.
  2. 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
  3. 3Click Add and start a new chat — Weaviate capabilities are ready to use.
Configuration · claude_desktop_config.jsonCopy
{
  "mcpServers": {
    "weaviate-mcp": {
      "url": "https://edge.vinkius.com/vk_preview_ewBpKaEj3GWOCB4HdPc1Tag8b8QBrbGJDpCpFWfW/mcp"
    }
  }
}
Copy into chat04

  • Find all documents in the 'research_papers' collection that discuss renewable energy trends.

  • Search for objects in 'customer_feedback' that sound like they are complaining about pricing.

  • List all objects in the 'product_catalog' to see what we have in stock.

  • Get the full details for the object with the ID 'user_12345' in the 'profiles' collection.

  • Claude
  • ChatGPT
  • Cursor
  • VS Code
  • Windsurf
  • Claude Code
  • JetBrains
  • Cline

Start here

Connect Weaviate once, then ask.

Just provide your connection details. Your credentials stay encrypted, and you can immediately start asking questions in your preferred AI client.

Connect Weaviate to your AI

FAQ

How this task behaves

  • 01

    Can the AI delete data from my Weaviate instance?

    No. This MCP only has tools to list, search, and get details. It cannot delete or modify your data.

  • 02

    Does the AI need to know my vector dimensions?

    No. The AI handles the vector math and the search process through the provided tools.

  • 03

    Can the AI update existing objects?

    No. There are no tools provided to update or write new data to Weaviate.

  • 04

    How does it find things if I don't use the exact words?

    It uses the search_near_vector tool to find objects based on mathematical similarity in meaning.

  • 05

    Can the AI create new collections?

    No. The AI can only interact with collections that already exist.

  • More questions about Weaviate? The Connector page answers them. See everything the Weaviate Connector can do

Connect Weaviate to Claude, Cursor, ChatGPT & more