Implementation:Googleapis Python genai GenerateContentResponse Access

Knowledge Sources	googleapis/python-genai Google Gen AI Python SDK
Domains	NLP, Data_Extraction
Last Updated	2026-02-15 00:00 GMT

Overview

Concrete tool for accessing generated text, parts, function calls, and metadata from model responses provided by the google-genai types module.

Description

GenerateContentResponse is the Pydantic model returned by generate_content and generate_content_stream. It provides convenience properties for common access patterns: .text returns concatenated text from the first candidate, .parts returns the raw parts list, .function_calls returns any function call requests from the model, and .usage_metadata provides token counting information. The .candidates field provides access to all response candidates with their safety ratings and finish reasons.

Usage

Access .text for simple text extraction. Use .function_calls in function-calling workflows to determine which functions the model wants to invoke. Use .usage_metadata for cost tracking and monitoring. Check .candidates[0].finish_reason to understand why generation stopped (STOP, MAX_TOKENS, SAFETY, etc.).

Code Reference

Source Location

• Repository: googleapis/python-genai
• File: google/genai/types.py
• Lines: L6781-6920

Signature

class GenerateContentResponse(_common.BaseModel):
    """Response from generate_content."""

    candidates: Optional[list[Candidate]] = None
    model_version: Optional[str] = None
    usage_metadata: Optional[GenerateContentResponseUsageMetadata] = None
    automatic_function_calling_history: Optional[list[Content]] = None

    @property
    def text(self) -> Optional[str]:
        """Concatenated text from all text parts in the first candidate."""

    @property
    def parts(self) -> Optional[list[Part]]:
        """Parts list from the first candidate's content."""

    @property
    def function_calls(self) -> Optional[list[FunctionCall]]:
        """Function calls from the first candidate."""

Import

from google.genai import types

I/O Contract

Inputs

Name	Type	Required	Description
(returned by generate_content)	GenerateContentResponse	N/A	This is a return type, not directly constructed by users

Outputs

Name	Type	Description
.text	Optional[str]	Concatenated text from all text parts in the first candidate
.parts	Optional[list[Part]]	Raw parts list from the first candidate
.function_calls	Optional[list[FunctionCall]]	Function call requests from the model
.candidates	Optional[list[Candidate]]	All response candidates with safety ratings and finish reasons
.usage_metadata	Optional[GenerateContentResponseUsageMetadata]	Token counts: prompt_token_count, candidates_token_count, total_token_count

Usage Examples

Basic Text Extraction

from google import genai

client = genai.Client(api_key="YOUR_API_KEY")

response = client.models.generate_content(
    model="gemini-2.0-flash",
    contents="What is Python?"
)

# Simple text access
print(response.text)

# Token usage
print(f"Prompt tokens: {response.usage_metadata.prompt_token_count}")
print(f"Response tokens: {response.usage_metadata.candidates_token_count}")
print(f"Total tokens: {response.usage_metadata.total_token_count}")

Checking Finish Reason

response = client.models.generate_content(
    model="gemini-2.0-flash",
    contents="Tell me a story."
)

# Check why generation stopped
candidate = response.candidates[0]
print(f"Finish reason: {candidate.finish_reason}")
# Possible values: STOP, MAX_TOKENS, SAFETY, RECITATION, OTHER

Streaming Response Processing

full_text = ""
for chunk in client.models.generate_content_stream(
    model="gemini-2.0-flash",
    contents="Write an essay about AI ethics."
):
    if chunk.text:
        full_text += chunk.text
        print(chunk.text, end="")

print(f"\nTotal length: {len(full_text)}")

Related Pages

Implements Principle

• Principle:Googleapis_Python_genai_Response_Processing