Rerank

Rerank documents with Google Vertex AI

Portkey exposes Google Vertex AI’s Discovery Engine ranking API through the unified /v1/rerank endpoint.

Enable the Discovery Engine API on the GCP project backing your Vertex AI provider before making rerank requests.

Supported Models

• semantic-ranker-default@latest (default)
• semantic-ranker-fast@latest
• semantic-ranker-512@latest
• Specific versions: semantic-ranker-default-004, semantic-ranker-fast-004, semantic-ranker-default-003, semantic-ranker-default-002

Rerank documents

cURL

curl --location 'https://api.portkey.ai/v1/rerank' \
--header 'x-portkey-api-key: $PORTKEY_API_KEY' \
--header 'x-portkey-provider: @VERTEX_AI_PROVIDER_SLUG' \
--header 'Content-Type: application/json' \
--data '{
    "model": "semantic-ranker-default@latest",
    "query": "What is the capital of the United States?",
    "top_n": 3,
    "documents": [
        "Carson City is the capital city of the American state of Nevada.",
        "The Commonwealth of the Northern Mariana Islands is a group of islands in the Pacific Ocean. Its capital is Saipan.",
        "Washington, D.C. is the capital of the United States.",
        "Capitalization is the use of a capital letter at the start of a word."
    ]
}'

from portkey_ai import Portkey

client = Portkey(
    api_key="PORTKEY_API_KEY",
    provider="@VERTEX_AI_PROVIDER_SLUG"
)

result = client.post(
    "/rerank",
    model="semantic-ranker-default@latest",
    query="What is the capital of the United States?",
    top_n=3,
    documents=[
        "Carson City is the capital city of the American state of Nevada.",
        "The Commonwealth of the Northern Mariana Islands is a group of islands in the Pacific Ocean. Its capital is Saipan.",
        "Washington, D.C. is the capital of the United States.",
        "Capitalization is the use of a capital letter at the start of a word."
    ]
)

print(result)

import Portkey from 'portkey-ai';

const portkey = new Portkey({
    apiKey: "PORTKEY_API_KEY",
    provider: "@VERTEX_AI_PROVIDER_SLUG"
});

const result = await portkey.post('/rerank', {
    model: "semantic-ranker-default@latest",
    query: "What is the capital of the United States?",
    top_n: 3,
    documents: [
        "Carson City is the capital city of the American state of Nevada.",
        "The Commonwealth of the Northern Mariana Islands is a group of islands in the Pacific Ocean. Its capital is Saipan.",
        "Washington, D.C. is the capital of the United States.",
        "Capitalization is the use of a capital letter at the start of a word."
    ]
});

console.log(result);

Parameters

Parameter	Type	Required	Description
model	string	Yes	Ranking model to use (see supported models)
query	string	Yes	Query against which documents are ranked
documents	string[] | object[]	Yes	Documents to rank. Accepts strings or { text, title? } objects (max 200)
top_n	integer	No	Number of top results to return. Omit to return all
return_documents	boolean	No	Set to false to receive only IDs and scores (default true)

Related Resources

Google Vertex AI Ranking API

Official documentation for the Vertex AI Discovery Engine ranking API.