Add a base_model_paths property to _DiffusionServingModels.

tests/entrypoints/openai_api/test_image_server.py

@@ -10,12 +10,14 @@
 import base64
 import io
 from argparse import Namespace
+from types import SimpleNamespace
 
 import pytest
 from fastapi.testclient import TestClient
 from PIL import Image
 from pytest_mock import MockerFixture
 from vllm import SamplingParams
+from vllm.entrypoints.openai.models.protocol import BaseModelPath
 
 from vllm_omni.entrypoints.openai.image_api_utils import (
     encode_image_base64,
@@ -243,6 +245,30 @@ def test_models_endpoint_no_engine():
     assert len(data["data"]) == 0
 
 
+def test_get_engine_and_model_reads_base_model_paths_in_diffusion_mode():
+    """Regression test for pure diffusion servers exposing base_model_paths."""
+    from vllm_omni.entrypoints.openai.api_server import (
+        _DiffusionServingModels,
+        _get_engine_and_model,
+    )
+
+    engine_client = SimpleNamespace(stage_list=["diffusion"])
+    state = SimpleNamespace(
+        engine_client=engine_client,
+        stage_configs=[{"stage_type": "diffusion"}],
+        openai_serving_models=_DiffusionServingModels(
+            [BaseModelPath(name="tencent/HunyuanImage-3.0-Instruct", model_path="/models/hunyuan")]
+        ),
+    )
+    raw_request = SimpleNamespace(app=SimpleNamespace(state=state))
+
+    returned_engine, model_name, stage_types = _get_engine_and_model(raw_request)
+
+    assert returned_engine is engine_client
+    assert model_name == "tencent/HunyuanImage-3.0-Instruct"
+    assert stage_types == ["diffusion"]
+
+
 def test_generate_single_image(test_client):
     """Test generating a single image"""
     response = test_client.post(

vllm_omni/entrypoints/openai/api_server.py

@@ -178,6 +178,10 @@ class _DiffusionServingModels:
     def __init__(self, base_model_paths: list[BaseModelPath]) -> None:
         self._base_model_paths = base_model_paths
 
+    @property
+    def base_model_paths(self) -> list[BaseModelPath]:
+        return self._base_model_paths
+
     async def show_available_models(self) -> ModelList:
         return ModelList(
             data=[