Use the Google GenAI SDK with Respan for observability

Overview

The Respan Google GenAI integration enables full observability for Gemini models through Google's official SDK. Route requests through the Respan gateway for automatic logging, monitoring, and analytics.

The integration supports all GenAI features including temperature and sampling controls, safety settings, system instructions, structured output via response schemas, and thinking configuration for reasoning models.

How it works

Install the Google GenAI SDK and initialize the client with your Respan API key and the Respan base URL (https://api.keywordsai.co/api/google/gemini).

Use the standard GenAI SDK methods - generate_content, streaming, and tool calls. Every request is logged automatically with token usage, cost, and latency metrics.

python

from google import genai

client = genai.Client(
    api_key="YOUR_RESPAN_API_KEY",
    http_options={"api_version": "v1beta", "base_url": "https://api.keywordsai.co/api/google/gemini"},
)

response = client.models.generate_content(
    model="gemini-2.5-flash",
    contents="Explain how transformers work",
)