• SEARCH
  • CLUSTER
  • REDUCE
  • HUMAN

How to create vector embeddings with your AI

Connect NVIDIA AI to your agent to turn raw text into mathematical vectors. Once linked, your AI handles the heavy lifting of dimensionality reduction.

Ask AI about this page

Short answer

How do I create vector embeddings with my AI?

You use the get_embeddings tool to transform text or documents into high-dimensional vectors. This allows your agent to perform semantic searches or group similar data points. Here is how the process works.

The results

What you get from vectorization

The mathematical outputs your AI generates from your data.

SEARCH

Semantic search capability

The AI uses the vectors to find information based on meaning rather than just keywords. You get results that actually match the intent of your query.

CLUSTER

Data grouping

Your agent can organize large datasets into logical clusters. This makes it easy to spot patterns in unstructured text.

REDUCE

Dimensionality management

The AI handles the complex math required to represent data in high-dimensional space. You receive clean, usable vectors.

HUMAN

Manual review needed

You still need to check the clusters or search results to ensure the mathematical similarity matches your real-world needs. The AI provides the math, but you provide the context.

The workflow

What your AI does when the data arrives.

The AI moves from raw input to usable mathematical representations.

  1. Data ingestion

    Your agent receives the text or document you want to process. It prepares the content for the embedding engine.

    get_embeddings

  2. Vector generation

    The AI calls the embedding tool to map the input into a high-dimensional vector space. This captures the semantic essence of the data.

    get_embeddings

  3. Dimensionality mapping

    The engine calculates the specific coordinates for each piece of data. This ensures every input has a unique mathematical fingerprint.

    get_embeddings

  4. Output delivery

    The AI returns the completed vectors to your client. You can then feed these into a vector database or use them for immediate analysis.

    get_embeddings

Try it

Copy these to start.

Use these prompts to trigger the embedding process.

Starting points

These are opening lines. Swap out the bracketed text with your actual file names or data sources.

NVIDIA Connector

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

Connector linkhttps://edge.vinkius.com/vk_preview_F8wZEFp9XAw3aowvQGCJgs0iHR9Eswli1t6PkJLP/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 — NVIDIA capabilities are ready to use.
Configuration · claude_desktop_config.jsonCopy
{
  "mcpServers": {
    "nvidia-ai-mcp": {
      "url": "https://edge.vinkius.com/vk_preview_F8wZEFp9XAw3aowvQGCJgs0iHR9Eswli1t6PkJLP/mcp"
    }
  }
}
Copy into chat04

  • Generate embeddings for the text in my research_notes.txt file.

  • Create vector embeddings for the customer feedback entries in this spreadsheet.

  • Convert the product descriptions in my catalog.csv into vectors for semantic search.

  • Get embeddings for the contents of the legal_documents folder so I can cluster them.

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

Start here

Connect NVIDIA AI once, then ask.

Just link your account once. Your credentials stay encrypted, and you can immediately start asking your AI to process data using NVIDIA's hardware.

Connect NVIDIA AI to your AI

FAQ

How this task behaves

  • 01

    Can the AI create embeddings for images too?

    No. This specific tool is designed for text-based data to create vector representations.

  • 02

    Will the AI delete my original files?

    No. The tool only reads the content to generate vectors. It does not have the capability to modify or delete your source files.

  • 03

    Do I need to manage the vector database myself?

    Yes. The AI generates the embeddings, but you are responsible for storing them in a vector database if you want to perform long-term searches.

  • 04

    Can the AI perform the actual search?

    The AI generates the vectors used for searching. The actual retrieval process depends on how you use those vectors in your specific application or database.

  • 05

    How does the AI know which text to embed?

    The AI only embeds the specific text or files you explicitly point to in your prompt.

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

Connect NVIDIA to Claude, Cursor, ChatGPT & more