All Products
Search
Document Center

OpenSearch:Multimodal embedding

Last Updated:Apr 01, 2026

Call the AI Search Open Platform multimodal embedding service to convert text and images into vectors for search and retrieval.

Prerequisites

Before you begin, ensure that you have:

  • Activated the AI Search Open Platform service. See Activate the service.

  • An API key for authentication. See Obtain an API key.

  • SDK version 2.1.0 or later. To install or upgrade, run:

    pip install --upgrade alibabacloud_searchplat20240529

Call the multimodal embedding service

The following example calls the multimodal embedding service with a text input.

from alibabacloud_tea_openapi.models import Config
from alibabacloud_searchplat20240529.client import Client
from alibabacloud_searchplat20240529.models import GetMultiModalEmbeddingRequest, GetMultiModalEmbeddingRequestInput

# Configure the client
config = Config(
    bearer_token="Replace with your API-KEY",
    endpoint="<your-api-endpoint>",
    protocol="http"
)
client = Client(config=config)

# Build the request
request = GetMultiModalEmbeddingRequest()
request.from_map({
    "input": [
        {"text": "Science and technology are the primary productive forces"}
    ]
})

# Call the service
# "default" is the workspace name; replace "ops-m2-encoder" with another supported service ID if needed
response = client.get_multi_modal_embedding("default", "ops-m2-encoder", request)
print(response)

Replace the following placeholders:

PlaceholderDescription
Replace with your API-KEYYour API key for authentication
<your-api-endpoint>Your API endpoint, without the http:// prefix

Parameters

The request body must not exceed 8 MB. For the full parameter reference, see Multimodal embedding API reference.

What's next