From dd64c2cc89577fb3fe666f51274375c6df0aa5bc Mon Sep 17 00:00:00 2001
From: Kushaan koul <kushaankoul@gmail.com>
Date: Sat, 5 Sep 2026 16:27:19 +0530
Subject: [PATCH] feat(proxy): support suppressing x-litellm-version header
 (#39113)

---
 litellm/__init__.py                           |  1 +
 litellm/proxy/common_request_processing.py    |  8 +++
 .../proxy/test_common_request_processing.py   | 58 +++++++++++++++++++
 3 files changed, 67 insertions(+)

diff --git a/litellm/__init__.py b/litellm/__init__.py
index 41a3789ab0d..89960583651 100644
--- a/litellm/__init__.py
+++ b/litellm/__init__.py
@@ -429,6 +429,7 @@ autorouter_presets_url: str = os.getenv(
     "https://raw.githubusercontent.com/BerriAI/litellm/main/litellm/proxy/public_endpoints/autorouter_presets.json",
 )
 suppress_debug_info: bool = False
+suppress_version_header: bool = False
 dynamodb_table_name: Optional[str] = None
 s3_callback_params: Optional[Dict] = None
 s3_audit_callback_params: Optional[Dict] = None
diff --git a/litellm/proxy/common_request_processing.py b/litellm/proxy/common_request_processing.py
index fc83c1ddeed..ec7581e6f40 100644
--- a/litellm/proxy/common_request_processing.py
+++ b/litellm/proxy/common_request_processing.py
@@ -1,3 +1,4 @@
+import os
 import asyncio
 import contextlib
 import json
@@ -1547,6 +1548,13 @@ class ProxyBaseLLMRequestProcessing:
     ) -> dict:
         exclude_values: Final = {"", None, "None"}
         hidden_params = hidden_params or {}
+
+        if (
+            getattr(litellm, "suppress_version_header", False)
+            or os.environ.get("LITELLM_SUPPRESS_VERSION_HEADER", "").lower() in ("true", "1")
+            or kwargs.pop("suppress_version_header", False)
+        ):
+            version = None
         timing_values: Final = _timing_values(
             hidden_params=hidden_params,
             logging_obj=litellm_logging_obj,
diff --git a/tests/test_litellm/proxy/test_common_request_processing.py b/tests/test_litellm/proxy/test_common_request_processing.py
index 6d6aad22ca3..5de3ec2aea2 100644
--- a/tests/test_litellm/proxy/test_common_request_processing.py
+++ b/tests/test_litellm/proxy/test_common_request_processing.py
@@ -544,6 +544,64 @@ class TestProxyBaseLLMRequestProcessing:
         assert headers["x-ratelimit-remaining-requests"] == "999"
         proxy_logging_obj.post_call_response_headers_hook.assert_awaited_once()
 
+    @pytest.mark.asyncio
+    async def test_build_litellm_proxy_success_headers_suppress_version(self):
+        """Test suppressing x-litellm-version header via global and env config."""
+        import litellm
+
+        mock_request = MagicMock()
+        mock_request.headers = {}
+
+        class _FakeGenaiResponse:
+            _hidden_params = {
+                "model_id": "deployment-model-id",
+                "cache_key": "ck-test",
+                "api_base": "https://generativelanguage.googleapis.com/v1beta",
+                "response_cost": 0.001,
+            }
+
+        mock_user = MagicMock()
+        mock_user.spend = 0.0
+        mock_user.tpm_limit = None
+        mock_user.rpm_limit = None
+        mock_user.max_budget = None
+        mock_user.allowed_model_region = None
+
+        logging_obj = MagicMock()
+        logging_obj.litellm_call_id = "call-id-test"
+        logging_obj.model_id = "deployment-model-id"
+        logging_obj.model_call_details = {"litellm_call_id": "call-id-test"}
+
+        proxy_logging_obj = MagicMock()
+        proxy_logging_obj.post_call_response_headers_hook = AsyncMock()
+
+        try:
+            litellm.suppress_version_header = True
+            headers = await ProxyBaseLLMRequestProcessing.build_litellm_proxy_success_headers_from_llm_response(
+                response=_FakeGenaiResponse(),
+                request_data={"model": "gemini/gemini-1.5-flash"},
+                request=mock_request,
+                user_api_key_dict=mock_user,
+                logging_obj=logging_obj,
+                version="9.9.9",
+                proxy_logging_obj=proxy_logging_obj,
+            )
+            assert "x-litellm-version" not in headers
+        finally:
+            litellm.suppress_version_header = False
+
+        headers_kw = await ProxyBaseLLMRequestProcessing.build_litellm_proxy_success_headers_from_llm_response(
+            response=_FakeGenaiResponse(),
+            request_data={"model": "gemini/gemini-1.5-flash"},
+            request=mock_request,
+            user_api_key_dict=mock_user,
+            logging_obj=logging_obj,
+            version="9.9.9",
+            proxy_logging_obj=proxy_logging_obj,
+            suppress_version_header=True,
+        )
+        assert "x-litellm-version" not in headers_kw
+
     @pytest.mark.asyncio
     async def test_build_litellm_proxy_success_headers_streaming_style_iterator(self):
         """AsyncGoogleGenAIGenerateContentStreamingIterator sets _hidden_params at init; headers must propagate."""