RerankResponse

Contains reranked document scores.

For the list of response object types, see Response Objects.

Use RerankResponse to inspect reranked items and to access scores through list-style indexing.

Attributes

Attribute	Type	Description
items	list of RerankResultItem	Reranked result items.

RerankResultItem

Represents one reranked result returned in the items list of a RerankResponse.

Attributes

Attribute	Type	Description
index	int	Zero-based index of the document in the input list.
score	int or float	Relevance score assigned by the reranking model.

Iterate Items

Iterate over response.items to inspect each reranked result.

response = client.rerank(
    model_name="reranker_model",
    query="best trail shoes",
    documents=["Trail running shoes with durable grip.", "Formal leather shoes."],
)

for item in response.items:
    print(item.index, item.score)

Methods

Method	Return Type	Description
len(response)	int	Returns the number of reranked result items.
response[index]	RerankResultItem	Returns the reranked result item at the specified zero-based index.
model_dump()	dict[str, Any]	Returns a Pydantic dictionary representation of the object.
model_dump_json()	str	Returns a Pydantic JSON string representation of the object.
to_dict()	dict[str, Any]	Returns a dictionary representation compatible with earlier SDK response handling.
to_json()	str	Returns a JSON string representation compatible with earlier SDK response handling.
to_str()	str	Returns a readable string representation compatible with earlier SDK response handling.

Iterate Attributes

Use serialization helpers to iterate response attributes and item attributes.

response = client.rerank(
    model_name="reranker_model",
    query="best trail shoes",
    documents=[
        "Trail running shoes with durable grip.",
        "Formal leather shoes.",
        "Waterproof hiking shoes.",
    ],
)

payload = response.to_dict()

for attribute, value in payload.items():
    print(attribute, value)

for item in response.items:
    item_payload = item.to_dict()
    for attribute, value in item_payload.items():
        print(attribute, value)

Sample Response

{
  "items": [
    {
      "index": 0,
      "score": 0.9821
    },
    {
      "index": 2,
      "score": 0.8114
    }
  ]
}

Returned By

Returned by: rerank().