POST
/
v1beta
/
models
/
{model}
:embedContent
curl -X POST "https://api.example.com/v1beta/models/text-embedding-004:embedContent?key=YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "models/text-embedding-004",
    "content": {
      "parts": [
        {
          "text": "What is the meaning of life?"
        }
      ]
    },
    "taskType": "RETRIEVAL_QUERY"
  }'

{
  "embedding": {
    "values": [
      0.013168523,
      -0.008711934,
      -0.046782676,
      0.000699 5056,
      -0.024563482,
      ...
    ]
  }
}
Official documentation: https://ai.google.dev/gemini-api/docs/embeddings
Generate vector embeddings for text using Google Gemini embedding models. Embeddings are useful for semantic search, clustering, recommendations, and other machine learning tasks.

​
Request Parameters

​
key
string
required
API key.
​
model
string
required
The embedding model to use (e.g., “models/text-embedding-004”).
​
content
object
required
Content to embed.
  • parts (array): Array of content parts
    • text (string): Text to embed
​
taskType
string
Task type for the embedding.Options:
  • RETRIEVAL_QUERY: Query in a search/retrieval setting
  • RETRIEVAL_DOCUMENT: Document in a search/retrieval setting
  • SEMANTIC_SIMILARITY: Semantic similarity task
  • CLASSIFICATION: Classification task
  • CLUSTERING: Clustering task
​
title
string
Optional title for the content (only for RETRIEVAL_DOCUMENT task type).

​
Response

Returns a vector embedding for the input text. 
curl -X POST "https://api.example.com/v1beta/models/text-embedding-004:embedContent?key=YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "models/text-embedding-004",
    "content": {
      "parts": [
        {
          "text": "What is the meaning of life?"
        }
      ]
    },
    "taskType": "RETRIEVAL_QUERY"
  }'

{
  "embedding": {
    "values": [
      0.013168523,
      -0.008711934,
      -0.046782676,
      0.000699 5056,
      -0.024563482,
      ...
    ]
  }
}

​
Batch Embeddings

You can also generate embeddings for multiple texts in a single request using the batchEmbedContents endpoint: 
POST /v1beta/models/{model}:batchEmbedContents
This allows you to embed multiple pieces of content efficiently in one API call.