Short timeouts lead to ERROR 500

[svilupp]

Update: The actual error is the fact that low timeouts return a confusing generic error (500).

Expected behavior: Setting the timeout via HttpOptions as illustrated in the docs should be working, eg, HttpOptions(timeout=1000).

Actual behavior: If I set a timeout (see example below), I get an immediate API response: Error: 500 INTERNAL. {'error': {'code': 500, 'message': 'An internal error has occurred. Please retry or report in https://developers.generativeai.google/guide/troubleshooting', 'status': 'INTERNAL'}}

Environment details

• Programming language: Python
• OS: MacOS Sequoia 15.6
• Language runtime version: 3.13
• Package version: 1.32

Steps to reproduce

1. Run the below script and you'll get Error: 500 INTERNAL. {'error': {'code': 500, 'message': 'An internal error has occurred. Please retry or report in https://developers.generativeai.google/guide/troubleshooting', 'status': 'INTERNAL'}}

IF you comment out the line setting timeout in the client, it all works.

MRE

#!/usr/bin/env python3
"""
Minimal script to test Google GenAI direct client calls with timeout.
"""
import asyncio
import os
from google.genai import Client
from google.genai.types import HttpOptions
from dotenv import load_dotenv

load_dotenv()

async def main():
    # Get API key from environment
    api_key = os.getenv("GEMINI_API_KEY")
    if not api_key:
        print("Error: GEMINI_API_KEY environment variable not set")
        return
    
    # Create client with timeout
    client = Client(
        api_key=api_key,
        http_options=HttpOptions(timeout=20.0*100),  # 2 seconds timeout
    )
    
    try:
        # Use the aio client for async operations
        response = await client.aio.models.generate_content(
            model="gemini-2.5-flash",
            contents="Write a 50 words long, detailed haiku about a rainy day in London. Include vivid imagery and emotions.",
        )
        
        # Extract text from response
        print(f"Response: {response.text}")
        
    except Exception as e:
        print(f"Error: {e}")
        print(f"Error type: {type(e)}")


if __name__ == "__main__":
    asyncio.run(main())

Thank you for any advice on how to avoid this :)

[DouweM]

The culprit appears to be the X-Server-Timeout header that's set here:

python-genai/google/genai/_api_client.py

Line 167 in cd2620c

headers['X-Server-Timeout'] = str(math.ceil(timeout_in_seconds))

Note that it works if you increase the timeout to actually be 20 seconds (20000, 20.0*100 is only 2000 = 2 seconds). It seems like there's a lower limit on the server side.

[svilupp]

My bad! thank you!

[DouweM]

@svilupp I think the issue is worth keeping open because the too-low value should be reported explicitly instead of through a 500 error

[janasangeetha]

Hi Users,
Thank you for reaching out!
We appreciate you taking the time to share your thoughts with us, and we'll take this to the internal team. We’ll keep you updated once we have more information.
Thank you!