diff --git a/litellm/proxy/batches_endpoints/endpoints.py b/litellm/proxy/batches_endpoints/endpoints.py
index 98492bcc2d..03b9ac3dea 100644
--- a/litellm/proxy/batches_endpoints/endpoints.py
+++ b/litellm/proxy/batches_endpoints/endpoints.py
@@ -31,7 +31,6 @@ from litellm.proxy.openai_files_endpoints.common_utils import (
     get_original_file_id,
     prepare_data_with_credentials,
 )
-
 from litellm.proxy.utils import handle_exception_on_proxy, is_known_model
 from litellm.types.llms.openai import LiteLLMBatchCreateRequest
 
@@ -112,7 +111,10 @@ async def create_batch(  # noqa: PLR0915
             is_router_model = is_known_model(model=router_model, llm_router=llm_router)
 
         custom_llm_provider = (
-            provider or data.pop("custom_llm_provider", None) or "openai"
+            provider
+            or data.pop("custom_llm_provider", None)
+            or get_custom_llm_provider_from_request_headers(request=request)
+            or "openai"
         )
         _create_batch_data = LiteLLMBatchCreateRequest(**data)
         input_file_id = _create_batch_data.get("input_file_id", None)
diff --git a/tests/openai_endpoints_tests/test_openai_batches_endpoint.py b/tests/openai_endpoints_tests/test_openai_batches_endpoint.py
index 5a84900f87..be931f2f4b 100644
--- a/tests/openai_endpoints_tests/test_openai_batches_endpoint.py
+++ b/tests/openai_endpoints_tests/test_openai_batches_endpoint.py
@@ -5,12 +5,13 @@ import asyncio
 import aiohttp, openai
 from openai import OpenAI, AsyncOpenAI
 from typing import Optional, List, Union
-from test_openai_files_endpoints import upload_file, delete_file
 import os
 import sys
 import time
 from unittest.mock import patch, MagicMock, AsyncMock
 
+from litellm.proxy.batches_endpoints.endpoints import create_batch
+
 
 BASE_URL = "http://localhost:4000"  # Replace with your actual base URL
 API_KEY = "sk-1234"  # Replace with your actual API key
@@ -222,6 +223,89 @@ def test_vertex_batches_endpoint():
     pass
 
 
+@pytest.mark.asyncio
+async def test_create_batch_respects_custom_llm_provider_header():
+    request_body = {
+        "input_file_id": "file-batch-123",
+        "endpoint": "/v1/chat/completions",
+        "completion_window": "24h",
+    }
+
+    response_object = MagicMock()
+    response_object._hidden_params = {}
+    response_object.id = "batch-id"
+    response_object.input_file_id = request_body["input_file_id"]
+
+    mock_user_api_key_dict = MagicMock(
+        tpm_limit=0,
+        rpm_limit=0,
+        max_budget=0,
+        spend=0,
+        allowed_model_region="",
+    )
+    mock_user_api_key_dict.metadata = {}
+
+    request = MagicMock()
+    request.headers = {"custom-llm-provider": "vertex_ai"}
+    request.query_params = {}
+    request.method = "POST"
+
+    fastapi_response = MagicMock()
+    fastapi_response.headers = {}
+
+    logging_obj = MagicMock()
+
+    mock_proxy_logging_obj = MagicMock()
+    mock_proxy_logging_obj.post_call_success_hook = AsyncMock(
+        return_value=response_object
+    )
+    mock_proxy_logging_obj.post_call_failure_hook = AsyncMock()
+    mock_proxy_logging_obj.update_request_status = AsyncMock()
+
+    with patch(
+        "litellm.proxy.batches_endpoints.endpoints._read_request_body",
+        new_callable=AsyncMock,
+    ) as mock_read_body, patch(
+        "litellm.proxy.batches_endpoints.endpoints.ProxyBaseLLMRequestProcessing.common_processing_pre_call_logic",
+        new_callable=AsyncMock,
+    ) as mock_common_processing, patch(
+        "litellm.proxy.batches_endpoints.endpoints.litellm.acreate_batch",
+        new_callable=AsyncMock,
+    ) as mock_acreate_batch, patch(
+        "litellm.proxy.proxy_server.general_settings",
+        new={},
+    ), patch(
+        "litellm.proxy.proxy_server.proxy_config",
+        new={},
+    ), patch(
+        "litellm.proxy.proxy_server.version",
+        new="test-version",
+    ), patch(
+        "litellm.proxy.proxy_server.llm_router",
+        new=None,
+    ), patch(
+        "litellm.proxy.proxy_server.proxy_logging_obj",
+        new=mock_proxy_logging_obj,
+    ), patch(
+        "litellm.proxy.batches_endpoints.endpoints.asyncio.create_task",
+        new=lambda coro: asyncio.ensure_future(coro),
+    ):
+        mock_read_body.return_value = dict(request_body)
+        mock_common_processing.return_value = (dict(request_body), logging_obj)
+        mock_acreate_batch.return_value = response_object
+
+        response = await create_batch(
+            request=request,
+            fastapi_response=fastapi_response,
+            user_api_key_dict=mock_user_api_key_dict,
+        )
+
+    mock_acreate_batch.assert_awaited_once()
+    called_kwargs = mock_acreate_batch.call_args.kwargs
+    assert called_kwargs["custom_llm_provider"] == "vertex_ai"
+    assert response is response_object
+
+
 @pytest.mark.skip(reason="Local only test to verify if things work well")
 @pytest.mark.asyncio
 async def test_list_batches_with_target_model_names():