fix(vllm): list hosted_vllm/* models without an api key

Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>

litellm/llms/vllm/common_utils.py

@@ -1,3 +1,4 @@
+from types import MappingProxyType
 from typing import Final
 
 import httpx
@@ -45,16 +46,17 @@ class VLLMModelInfo(BaseLLMModelInfo):
 
     @staticmethod
     def get_api_base(api_base: str | None = None) -> str | None:
-        api_base = api_base or get_secret_str("VLLM_API_BASE")
+        api_base = api_base or get_secret_str("HOSTED_VLLM_API_BASE") or get_secret_str("VLLM_API_BASE")
         if api_base is None:
             raise ValueError(
-                "VLLM_API_BASE is not set. Please set the environment variable, to use VLLM's pass-through - `{LITELLM_API_BASE}/vllm/{endpoint}`."
+                "HOSTED_VLLM_API_BASE or VLLM_API_BASE is not set. Please set an environment variable to use "
+                "VLLM's pass-through - `{LITELLM_API_BASE}/vllm/{endpoint}`."
             )
         return api_base
 
     @staticmethod
     def get_api_key(api_key: str | None = None) -> str | None:
-        return None
+        return api_key or get_secret_str("HOSTED_VLLM_API_KEY")
 
     @staticmethod
     def get_base_model(model: str) -> str | None:
@@ -64,14 +66,17 @@ class VLLMModelInfo(BaseLLMModelInfo):
         api_base = VLLMModelInfo.get_api_base(api_base)
         api_key = VLLMModelInfo.get_api_key(api_key)
         endpoint: Final = "/v1/models"
-        if api_base is None or api_key is None:
+        if api_base is None:
             raise ValueError(
-                "VLLM_API_BASE or VLLM_API_KEY is not set. Please set the environment variable, to query VLLM's `/models` endpoint."
+                "HOSTED_VLLM_API_BASE or VLLM_API_BASE is not set. Please set an environment variable to query "
+                "VLLM's `/models` endpoint."
             )
 
         url: Final = add_path_to_api_base(api_base, endpoint)
+        headers: Final = MappingProxyType({"Authorization": f"Bearer {api_key}"}) if api_key is not None else None
         response: Final = litellm.module_level_client.get(
             url=url,
+            headers=headers,
         )
 
         response.raise_for_status()

tests/unit/llms/vllm/test_common_utils.py

@@ -0,0 +1,52 @@
+import httpx
+import pytest
+import respx
+
+from litellm.llms.vllm.common_utils import VLLMModelInfo
+
+MODELS_URL = "http://vllm.example/v1/models"
+MODEL_RESPONSE = {"data": [{"id": "Qwen/Qwen3-8B"}]}
+
+
+def test_get_models_without_api_key_sends_no_authorization_header(monkeypatch: pytest.MonkeyPatch) -> None:
+    monkeypatch.delenv("HOSTED_VLLM_API_KEY", raising=False)
+
+    with respx.mock:
+        route = respx.get(MODELS_URL).mock(return_value=httpx.Response(200, json=MODEL_RESPONSE))
+        models = VLLMModelInfo().get_models(api_base="http://vllm.example/v1")
+
+    assert models == ["Qwen/Qwen3-8B"]
+    assert "authorization" not in route.calls[0].request.headers
+
+
+def test_get_models_sends_api_key_as_bearer_token() -> None:
+    with respx.mock:
+        route = respx.get(MODELS_URL).mock(return_value=httpx.Response(200, json=MODEL_RESPONSE))
+        models = VLLMModelInfo().get_models(
+            api_key="vllm-test-key",
+            api_base="http://vllm.example/v1",
+        )
+
+    assert models == ["Qwen/Qwen3-8B"]
+    assert route.calls[0].request.headers["authorization"] == "Bearer vllm-test-key"
+
+
+def test_get_models_requires_api_base(monkeypatch: pytest.MonkeyPatch) -> None:
+    monkeypatch.delenv("HOSTED_VLLM_API_BASE", raising=False)
+    monkeypatch.delenv("VLLM_API_BASE", raising=False)
+
+    with pytest.raises(ValueError, match="HOSTED_VLLM_API_BASE"):
+        VLLMModelInfo().get_models(api_key="vllm-test-key")
+
+
+def test_get_api_key_falls_back_to_hosted_vllm_environment_variable(monkeypatch: pytest.MonkeyPatch) -> None:
+    monkeypatch.setenv("HOSTED_VLLM_API_KEY", "hosted-vllm-test-key")
+
+    assert VLLMModelInfo.get_api_key() == "hosted-vllm-test-key"
+
+
+def test_get_api_base_prefers_hosted_vllm_environment_variable(monkeypatch: pytest.MonkeyPatch) -> None:
+    monkeypatch.setenv("HOSTED_VLLM_API_BASE", "http://hosted-vllm.example/v1")
+    monkeypatch.setenv("VLLM_API_BASE", "http://vllm.example/v1")
+
+    assert VLLMModelInfo.get_api_base() == "http://hosted-vllm.example/v1"