What is embedding generation?

Embedding generation turns text into a numeric vector that machines can compare. That is useful for semantic search, clustering, recommendations, and similarity checks.

Tool description

This tool sends your text to a selected embedding model and shows the resulting vector and its dimensions. It supports OpenAI, Google, and Voyage AI models, so you can compare providers or test an embedding setup without leaving the app.

Example

Input:

Reset your password from the account settings page.

Output:

{
  "model": "openai/text-embedding-3-small",
  "dimensions": 1536,
  "embedding": [0.0123, -0.0441, 0.0087]
}

Features

  • Lets you choose between supported embedding models from OpenAI, Google, and Voyage AI
  • Shows the embedding dimension and the raw vector as JSON
  • Downloads the result as embedding.json

How it works

The tool trims the input text, sends it with the selected model to the embedding API, and renders the returned vector in the output area. If the model changes, the vector shape can change too, so the dimensions field helps you confirm what you got back.

Tips

  • Use short, representative text when you are comparing models.
  • Different providers and models often return vectors with different dimensions.
  • This is a good way to sanity check a semantic search pipeline before you wire it into code.