diff --git a/litellm/llms/custom_httpx/llm_http_handler.py b/litellm/llms/custom_httpx/llm_http_handler.py
index c9ab3c648acd..7939e2626f16 100644
--- a/litellm/llms/custom_httpx/llm_http_handler.py
+++ b/litellm/llms/custom_httpx/llm_http_handler.py
@@ -2026,6 +2026,7 @@ async def async_anthropic_messages_handler(
             litellm_params={
                 "preset_cache_key": None,
                 "stream_response": {},
+                "model_info": kwargs.get("model_info"),
                 **anthropic_messages_optional_request_params,
             },
             custom_llm_provider=custom_llm_provider,
diff --git a/tests/test_litellm/llms/custom_httpx/test_llm_http_handler.py b/tests/test_litellm/llms/custom_httpx/test_llm_http_handler.py
index 07a61c9c1049..083f4a97ab62 100644
--- a/tests/test_litellm/llms/custom_httpx/test_llm_http_handler.py
+++ b/tests/test_litellm/llms/custom_httpx/test_llm_http_handler.py
@@ -334,6 +334,79 @@ async def test_async_anthropic_messages_handler_passes_litellm_metadata():
     assert kwargs_arg["litellm_metadata"]["model_info"] == custom_model_info
 
 
+@pytest.mark.asyncio
+async def test_async_anthropic_messages_handler_forwards_router_model_info():
+    """Ensure router deployment model_info is forwarded into litellm_params.
+
+    The Router stamps kwargs['model_info'] on every deployment dispatch via
+    _update_kwargs_with_deployment. Downstream cooldown / success callbacks
+    (router.deployment_callback_on_failure, deployment_callback_on_success)
+    look up the deployment id via kwargs['litellm_params']['model_info']['id'].
+    If async_anthropic_messages_handler builds its own litellm_params dict
+    without forwarding model_info, the id is missing and cooldown is silently
+    skipped for /v1/messages requests under the Router.
+    """
+    handler = BaseLLMHTTPHandler()
+
+    mock_config = Mock()
+    mock_config.validate_anthropic_messages_environment = Mock(
+        return_value=({"x-api-key": "test-key"}, "https://api.anthropic.com")
+    )
+    mock_config.transform_anthropic_messages_request = Mock(
+        return_value={"model": "claude-sonnet-4-20250514", "messages": []}
+    )
+
+    mock_client = AsyncMock()
+    mock_response = Mock()
+    mock_response.status_code = 200
+    mock_response.json.return_value = {
+        "id": "msg_123",
+        "type": "message",
+        "role": "assistant",
+        "content": [{"type": "text", "text": "Hello!"}],
+        "model": "claude-sonnet-4-20250514",
+        "stop_reason": "end_turn",
+    }
+    mock_client.post = AsyncMock(return_value=mock_response)
+
+    mock_logging_obj = Mock()
+    mock_logging_obj.update_from_kwargs = Mock()
+    mock_logging_obj.model_call_details = {}
+    mock_logging_obj.stream = False
+
+    deployment_model_info = {
+        "id": "deployment-123",
+        "db_model": False,
+    }
+
+    try:
+        await handler.async_anthropic_messages_handler(
+            model="claude-sonnet-4-20250514",
+            messages=[{"role": "user", "content": "Hello"}],
+            anthropic_messages_provider_config=mock_config,
+            anthropic_messages_optional_request_params={},
+            custom_llm_provider="anthropic",
+            litellm_params=GenericLiteLLMParams(),
+            logging_obj=mock_logging_obj,
+            client=mock_client,
+            kwargs={"model_info": deployment_model_info},
+        )
+    except Exception:
+        pass
+
+    mock_logging_obj.update_from_kwargs.assert_called_once()
+    call_kwargs = mock_logging_obj.update_from_kwargs.call_args
+    litellm_params_arg = (
+        call_kwargs.kwargs.get(
+            "litellm_params", call_kwargs[1].get("litellm_params", {})
+        )
+        if call_kwargs.kwargs
+        else call_kwargs[1].get("litellm_params", {})
+    )
+
+    assert litellm_params_arg.get("model_info") == deployment_model_info
+
+
 @pytest.mark.asyncio
 async def test_async_anthropic_messages_handler_header_priority():
     """