Implementation:Openai Openai python Response Output Text

Knowledge Sources	Openai_Openai_python OpenAI API Reference
Domains	API_Types, Responses_API
Last Updated	2026-02-15 00:00 GMT

Overview

Concrete type for representing text output content from the model along with citation annotations and log probabilities, provided by the openai-python SDK.

Description

ResponseOutputText is a Pydantic model representing a text output from the model. It contains the text content, a list of annotations (citations and file paths), and an optional logprobs field with token-level log probabilities. The type field is always "output_text".

The module also defines several supporting types:

• Annotation - A discriminated union TypeAlias of citation types: AnnotationFileCitation (file citations with file_id, filename, index), AnnotationURLCitation (web citations with start_index, end_index, title, url), AnnotationContainerFileCitation (container file citations with container_id, file_id, filename, start_index, end_index), and AnnotationFilePath (file paths with file_id, index).
• Logprob - Token log probability with token, bytes, logprob, and top_logprobs fields.
• LogprobTopLogprob - A top log probability candidate for a given token position.

Usage

Import this type when you need to inspect the text content, citations, or log probabilities within output message content items from the Responses API.

Code Reference

Source Location

• Repository: openai-python
• File: src/openai/types/responses/response_output_text.py

Signature

class ResponseOutputText(BaseModel):
    """A text output from the model."""

    annotations: List[Annotation]
    text: str
    type: Literal["output_text"]
    logprobs: Optional[List[Logprob]] = None

Import

from openai.types.responses import ResponseOutputText

I/O Contract

Fields

Name	Type	Required	Description
annotations	List[Annotation]	Yes	The annotations of the text output (citations, file paths).
text	str	Yes	The text output from the model.
type	Literal["output_text"]	Yes	The type of the output text. Always output_text.
logprobs	Optional[List[Logprob]]	No	Token-level log probabilities, if requested.

Annotation Union Members

Type	Discriminator Value	Key Fields
AnnotationFileCitation	"file_citation"	file_id, filename, index
AnnotationURLCitation	"url_citation"	start_index, end_index, title, url
AnnotationContainerFileCitation	"container_file_citation"	container_id, file_id, filename, start_index, end_index
AnnotationFilePath	"file_path"	file_id, index

Logprob Fields

Name	Type	Description
token	str	The token string.
bytes	List[int]	The byte representation of the token.
logprob	float	The log probability of the token.
top_logprobs	List[LogprobTopLogprob]	The top log probability candidates.

Usage Examples

import openai

client = openai.OpenAI()

response = client.responses.create(
    model="gpt-4o",
    input="What is the capital of France?",
    tools=[{"type": "web_search_preview"}],
)

for item in response.output:
    if item.type == "message":
        for content in item.content:
            if content.type == "output_text":
                print(content.text)
                # Inspect annotations (citations)
                for annotation in content.annotations:
                    if annotation.type == "url_citation":
                        print(f"  Citation: {annotation.title} - {annotation.url}")
                    elif annotation.type == "file_citation":
                        print(f"  File: {annotation.filename} ({annotation.file_id})")

Related Pages

• Environment:Openai_Openai_python_Python_3_9_Plus