12sec Latency for 2.5-flash-lite

Gemini API
,
1

A simple message `hi` has a 12 second latency when using Vertex AI

response = client.models.generate_content(
            model="gemini-2.5-flash-lite",
            contents="hi",
            config=GenerateContentConfig(
                response_mime_type="application/json",
                max_output_tokens=20,
                temperature=0,
                thinking_config=ThinkingConfig(
                    thinking_budget=0,
                ),
            ),
        )

This is making it completely unusable