Unable to obtain grounding data from response. using "gemini-3.1-flash-lite-preview"

[satounity]

(I don't know English so please excuse the machine translation.)

Grounding using Vertex AI Search in "gemini-3.1-flash-lite-preview".
No grounding data found in response.candidates[0].grounding_metadata.grounding_chunks.

[Example]
response = client.models.generate_content(...
...

print(response.candidates[0].grounding_metadata.grounding_chunks)
->No results

[satounity]

Add a note

――――――――――――――――――――
# model_id = "gemini-3-flash-preview"
model_id = "gemini-3.1-flash-lite-preview"
response = client.models.generate_content(
    model=model_id,
    contents=prompt_text,
    config=GenerateContentConfig(
        response_mime_type="application/json",
        # Grounding [Vertex AI Search]
        tools=[
            Tool(
                retrieval=Retrieval(
                    vertex_ai_search=VertexAISearch(
                        datastore=datastore,
                    )
                )
            )
        ],
        thinking_config=ThinkingConfig(
            thinking_level="LOW",
        ),
    ),
)
print(response.candidates[0].grounding_metadata.grounding_chunks)
――――――――――――――――――――

In the above code snippet,

●If the model is "gemini-3-flash-preview",
the
"print(response.candidates[0].grounding_metadata.grounding_chunks)"
command returns a result.

●If the model is "gemini-3.1-flash-lite-preview",
the
"print(response.candidates[0].grounding_metadata.grounding_chunks)"
command does not return a result.