fix: filter thought parts client-side when include_thoughts=False on Vertex AI

[nemetron]

Problem

On Vertex AI, ThinkingConfig(include_thoughts=False) does not suppress thought parts from the API response for image generation models (e.g. gemini-3.1-flash-image-preview). All intermediate thought text and draft images are returned in the response despite the config, making the flag a no-op on Vertex AI.

The part.thought boolean IS correctly set to True on thought parts by the API, so the information needed for client-side filtering is available.

Fixes #2239

Root Cause

The Vertex AI backend does not honor include_thoughts=False for this model family — it returns all thought parts regardless. The SDK passed them through without filtering.

Fix

Added _filter_thought_parts() in models.py, called in all four response paths (sync/async, unary/streaming). It inspects ThinkingConfig.include_thoughts and removes parts where part.thought is True when the value is explicitly False.

The filter is a strict opt-in:

• include_thoughts=False → thought parts stripped
• include_thoughts=True → unchanged
• include_thoughts=None / no ThinkingConfig / no config → unchanged

Tests

7 new unit tests covering all branches of _filter_thought_parts:

PASSED test_include_thoughts_false_removes_thought_parts
PASSED test_include_thoughts_true_preserves_all_parts
PASSED test_include_thoughts_none_preserves_all_parts
PASSED test_no_thinking_config_preserves_all_parts
PASSED test_no_config_preserves_all_parts
PASSED test_empty_candidates_is_safe
PASSED test_no_thought_parts_in_response

Reproduction (from issue)

from google import genai
from google.genai import types

client = genai.Client(vertexai=True, project="YOUR_PROJECT", location="us-central1")

response = client.models.generate_content(
    model="gemini-3.1-flash-image-preview",
    contents="Draw a red car",
    config=types.GenerateContentConfig(
        response_modalities=["TEXT", "IMAGE"],
        thinking_config=types.ThinkingConfig(include_thoughts=False),
    ),
)

for i, part in enumerate(response.candidates[0].content.parts):
    print(f"[{i}] thought={part.thought}, has_text={bool(part.text)}, has_image={bool(part.inline_data)}")
# After fix: only the final image part is returned
# [0] thought=None, has_text=False, has_image=True

[google-cla]

Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

View this failed invocation of the CLA check for more information.

For the most up to date status, view the checks section at the bottom of the pull request.

[Venkaiahbabuneelam]

Hi @nemetron, Thanks for reaching out us!

I noticed some checks have failed. Kindly resolve the conflicts.

Thanks