gemini-embed

Gemini Embeddings API

Generate text embeddings using Google Gemini API via REST.

Prerequisites

• Environment variable GOOGLE_API_KEY must be set
• API endpoint: https://generativelanguage.googleapis.com/v1beta
• Model: gemini-embedding-001

Workflow

Phase 1: Determine Embedding Type

• Single Embedding: For one text input
• Batch Embedding: For multiple texts (more efficient)

Phase 2: Configure Task Type (Optional)

Choose based on use case:

• RETRIEVAL_QUERY: For search queries
• RETRIEVAL_DOCUMENT: For documents to be searched
• SEMANTIC_SIMILARITY: For comparing text similarity
• CLASSIFICATION: For text classification
• CLUSTERING: For grouping similar texts

Phase 3: Execute API Call

---

1. Single Text Embedding

Basic Embedding

curl "https://generativelanguage.googleapis.com/v1beta/models/gemini-embedding-001:embedContent?key=$GOOGLE_API_KEY" \
    -H 'Content-Type: application/json' \
    -d '{
      "model": "models/gemini-embedding-001",
      "content": {
        "parts": [{"text": "Hello world"}]
      }
    }'

With Task Type

curl "https://generativelanguage.googleapis.com/v1beta/models/gemini-embedding-001:embedContent?key=$GOOGLE_API_KEY" \
    -H 'Content-Type: application/json' \
    -d '{
      "model": "models/gemini-embedding-001",
      "content": {
        "parts": [{"text": "What is machine learning?"}]
      },
      "task_type": "RETRIEVAL_QUERY"
    }'

With Output Dimensionality Control

Truncate embeddings to a smaller size for efficiency:

curl "https://generativelanguage.googleapis.com/v1beta/models/gemini-embedding-001:embedContent?key=$GOOGLE_API_KEY" \
    -H 'Content-Type: application/json' \
    -d '{
      "model": "models/gemini-embedding-001",
      "content": {
        "parts": [{"text": "Hello world"}]
      },
      "output_dimensionality": 256
    }'

---

2. Batch Embedding

Process multiple texts in a single API call: