Skip to content

Embeddings API

The Embeddings API converts text into vector embeddings for semantic search and similarity comparisons.

Overview

Generate vector embeddings for text using Unique AI's embedding models.

Methods

unique_sdk.Embeddings.create - Generate vector embeddings

Convert text strings into vector embeddings.

Parameters:

  • user_id (required)
  • company_id (required)
  • texts (required) - Array of strings to embed

Example:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
result = unique_sdk.Embeddings.create(
    user_id=user_id,
    company_id=company_id,
    texts=["hello", "world", "embeddings"]
)

# Access embeddings
for i, embedding in enumerate(result.embeddings):
    print(f"Text {i}: {len(embedding)} dimensions")
    print(f"First 5 values: {embedding[:5]}")

Response Format

1
2
3
4
5
6
7
{
    "embeddings": [
        [0.123, -0.456, 0.789, ...],  # Vector for first text
        [0.234, -0.567, 0.890, ...],  # Vector for second text
        ...
    ]
}

Each embedding is a list of floating-point numbers representing the text in vector space.