[Bug]: Reasoning content is missing when calling vertex_ai gemini 2.5 pro model from docker proxy server

[alexyzhou]

What happened?

Hi team,

I am deploying litellm/litellm docker image and added gemini 2.5 pro model with vertex_ai credentials. When I am trying to call chat/completions with this model, it seems the reasoning content will not return, and it will stuck for a while and return only the final response part.

The image I am deploying is: https://hub.docker.com/r/litellm/litellm
version: [v1.80.7.dev.3]

The code I am using to call litellm proxy server:

import requests

API_KEY = "YOUR_API_KEY"
BASE_URL = "LITELLM_DEPLOYMENG_ENDPOINT"
MODEL = "gemini-2.5-pro"

url = f"{BASE_URL}/v1/chat/completions"

headers = {
    "Authorization": f"Bearer {API_KEY}",
    "Content-Type": "application/json",
}

payload = {
    "model": MODEL,
    "stream": True,
    "messages": [
        {"role": "user", "content": "How many 'r' in 'cherry'?"},
    ],
}

with requests.post(url, headers=headers, json=payload, stream=True) as resp:
    resp.raise_for_status()

    for line in resp.iter_lines(decode_unicode=True):
        if line:
            print(line)

Relevant log output

Are you a ML Ops Team?

No

What LiteLLM version are you on ?

v1.80.7.dev.3

Twitter / LinkedIn details

No response