Skip to main content

Generate Embedding

Generate vector embeddings from text using OpenAI embedding models for semantic search and similarity matching.

Common Properties​

  • Name - The custom name of the node.
  • Color - The custom color of the node.
  • Delay Before (sec) - Waits in seconds before executing the node.
  • Delay After (sec) - Waits in seconds after executing node.
  • Continue On Error - Automation will continue regardless of any error. Default: false.

Inputs​

  • Connection Id - Connection identifier from Connect node.
  • Text - Text to convert into a vector embedding.
  • Use Robomotion AI Credits - Use Robomotion credits instead of your own API key.

Options​

  • Model - Embedding model:
    • text-embedding-3-small - Fast, cost-effective (default)
    • text-embedding-3-large - Higher accuracy
    • text-embedding-ada-002 - Legacy model
    • Custom - Specify custom model name
  • Custom Model - Custom model name (when Model is "Custom").
  • Dimensions - Output embedding dimensions (for v3 models). Lower dimensions = smaller, faster.
  • User - Unique end-user identifier.
  • Include Raw Response - Include usage info. Default: false.

Outputs​

  • Embedding - Vector embedding as an array of floating-point numbers.
  • Raw Response - Full response with usage info (when enabled).

How It Works​

Converts text into a numerical vector representation:

  1. Validates connection and input text
  2. Sends text to embedding model
  3. Receives vector representation
  4. Returns embedding array

Usage Examples​

Example 1: Embed Search Query​

Input:
- Text: "How to reset password"
- Model: text-embedding-3-small

Output:
- Embedding: [0.0234, -0.0156, 0.0891, ... ] (1536 dimensions)

Example 2: Reduced Dimensions​

Input:
- Text: "Product description for blue shirt"
- Model: text-embedding-3-small
- Dimensions: 512

Output:
- Embedding: [0.0234, -0.0156, ... ] (512 dimensions)

Example 3: High-Accuracy Embedding​

Input:
- Text: "Machine learning model evaluation metrics"
- Model: text-embedding-3-large

Output:
- Embedding: [0.0234, -0.0156, ... ] (3072 dimensions)

Requirements​

  • Connection Id from Connect node
  • Non-empty text input

Use Cases​

  • Semantic Search: Find similar documents based on meaning, not just keywords
  • Recommendation Systems: Find related products, articles, or content
  • Clustering: Group similar items together
  • Classification: Categorize text based on similarity to examples
  • Duplicate Detection: Find similar or duplicate content
  • Question Answering: Match questions to relevant answers

Tips for RPA Developers​

  • Model Selection: Use text-embedding-3-small for most cases. Use 3-large when accuracy is critical.
  • Dimensions: Reducing dimensions saves storage and computation with minimal accuracy loss.
  • Normalization: Embeddings are already normalized - ready for cosine similarity.
  • Batch Processing: Use Generate Batch Embeddings for multiple texts to save time and cost.
  • Storage: Store embeddings in databases with vector search (PostgreSQL pgvector, Pinecone, etc.).
  • Similarity: Use cosine similarity or the Similarity node to compare embeddings.

Common Errors​

"Text cannot be empty"

  • Provide text content to embed

"Custom Model cannot be empty"

  • Specify a model name when using Custom option