400 BadRequestError with Gemini provider – Invalid JSON payload ("metadata" and "store" fields)

[Mubashar-Bashir]

Please read this first

• Have you read the custom model provider docs, including the 'Common issues' section? ✅ Yes
• Have you searched for related issues? ✅ Yes

---

Describe the question

I'm using the OpenAI Agents SDK with a custom model provider (Google Gemini via https://generativelanguage.googleapis.com/v1beta/openai/) and encountering a BadRequestError.

The Gemini API rejects the request with this error:

BadRequestError: Error code: 400 - [{'error': {'code': 400, 'message': 'Invalid JSON payload received. Unknown name "metadata": Cannot find field.\nInvalid JSON payload received. Unknown name "store": Cannot find field.'}}]

It appears that the Agents SDK is sending metadata and store fields automatically, but the Gemini API doesn't support these.

---

Debug information

• Agents SDK version: v0.0.3
• Python version: 3.11
• Custom model provider: Google Gemini (Flash 1.5)/(Flash 2.0)
• Request Type: chat.completions.create

---

Repro steps

Here’s a minimal script to reproduce the issue:

from openai import AsyncOpenAI
from agents import Agent, Runner, OpenAIChatCompletionsModel
from agents.run import RunConfig
import nest_asyncio

nest_asyncio.apply()

# External Async client for custom provider
client = AsyncOpenAI(
    api_key=google_api_key,
    base_url="https://generativelanguage.googleapis.com/v1beta/openai/",
)

# Wrap client in SDK model
model = OpenAIChatCompletionsModel(
    model="gemini-1.5-flash",  # "gemini-2.0-flash"  
    openai_client=client,
)

config = RunConfig(
    model=model,
    model_provider=client,
    tracing_disabled=True
)

agent = Agent(name="Assistant", instructions="You are a helpful assistant")

# Run agent synchronously
result = Runner.run_sync(agent, "Write a haiku about recursion in programming.")
print(result.final_output)

**Metadata**
Please clarify:

Is there a way to prevent the SDK from sending unsupported fields like metadata or store?

Will future versions of the SDK provide better integration options for non-OpenAI endpoints?

[hamzaadil56]

Yes, I'm also facing the same bug. OpenAI has updated its library, which is causing some errors when providing Gemini Client.

I've worked with version 0.0.4, and in that, it was working fine. But when I started a new project with the latest version, it is causing issues when providing with Gemini Provider.

[Qureshihasaan]

Yes, facing the same issue of json payload....Try different things but can't debug it, Maybe, something new Package/Library is added to new version.

[shamoon-ahmed]

I'm having the same BadRequestError. How do I resolve it?

[DevHammad0]

I tested older versions (0.0.7, 0.0.6, 0.0.5, and 0.0.4) and they work fine without the metadata and store payload errors.

So it seems that version 0.0.8 is introducing this issue, possibly by including unsupported fields like metadata and store in the request payload, which are not supported by the Gemini API via the OpenAI-compatible endpoint (https://generativelanguage.googleapis.com/v1beta/openai/).

[shamoon-ahmed]

Yessir thanks a lot! I just tried the 0.0.4 version and it worked fine! Hope they resolve this issue in the latest versions as well

[Mrkhan9914626]

same issue

[jhammarstedt]

+1 on this

[exiao]

same issue with 0.0.8, breaking change. #431

@drewy-openai , @rm-openai

[BasedLukas]

same issue

[RahulVerma989]

I am also facing the same issue with groq and google openai compatible apis. I think we can easily fix it by not passing the metadata when it is null.

[rm-openai]

Sorry about that, fixing now.

[RahulVerma989]

For metadata property issue. I have added two review comments in openai_chatcompletions.py and openai_responses.py.

In short the changes will be:

current:
metadata=model_settings.metadata

change to:
metadata=self._non_null_or_not_given(model_settings.metadata)

[rm-openai]

fix in #455 is merged, I'll cut a new release once #456 merges

[rm-openai]

v0.0.9 should be out in a few mins. Let me know if you have any issues with it.

[rm-openai]

Closing. Please create a new issue for any follow ups!