[Refactor] Relocate endpoint tests to mirror serving code directory structure (#37504)

Signed-off-by: sfeng33 <4florafeng@gmail.com>

.buildkite/scripts/hardware_ci/run-amd-test.sh

@@ -336,11 +336,10 @@ apply_rocm_test_overrides() {
     --ignore=entrypoints/openai/chat_completion/test_audio.py \
     --ignore=entrypoints/openai/completion/test_shutdown.py \
     --ignore=entrypoints/openai/test_completion.py \
-    --ignore=entrypoints/openai/test_models.py \
+    --ignore=entrypoints/openai/models/test_models.py \
     --ignore=entrypoints/openai/test_lora_adapters.py \
     --ignore=entrypoints/openai/test_return_tokens_as_ids.py \
     --ignore=entrypoints/openai/chat_completion/test_root_path.py \
-    --ignore=entrypoints/openai/test_tokenization.py \
     --ignore=entrypoints/openai/completion/test_prompt_validation.py "}
   fi
 

pyproject.toml

@@ -121,7 +121,7 @@ python = "./.venv"
 # these files may be written in non english words
 extend-exclude = ["tests/models/fixtures/*", "tests/prompts/*", "tests/tokenizers_/*",
     "benchmarks/sonnet.txt", "tests/lora/data/*", "examples/pooling/token_embed/*", "build/*",
-    "vllm/third_party/*", "vllm/entrypoints/serve/instrumentator/static/*", "tests/entrypoints/openai/test_transcription_validation.py",
+    "vllm/third_party/*", "vllm/entrypoints/serve/instrumentator/static/*", "tests/entrypoints/openai/speech_to_text/test_transcription_validation.py",
     "docs/governance/process.md", "tests/v1/engine/test_fast_incdec_prefix_err.py", ".git/*"]
 ignore-hidden = false
 

tests/entrypoints/openai/test_messages.py → tests/entrypoints/anthropic/test_messages.py

@@ -5,7 +5,7 @@ import anthropic
 import pytest
 import pytest_asyncio
 
-from ...utils import RemoteOpenAIServer
+from tests.utils import RemoteOpenAIServer
 
 MODEL_NAME = "Qwen/Qwen3-0.6B"
 

tests/entrypoints/openai/test_models.py → tests/entrypoints/openai/models/test_models.py

@@ -5,7 +5,7 @@ import openai  # use the official client for correctness check
 import pytest
 import pytest_asyncio
 
-from ...utils import RemoteOpenAIServer
+from tests.utils import RemoteOpenAIServer
 
 # any model with a chat template should work here
 MODEL_NAME = "Qwen/Qwen3-0.6B"

tests/entrypoints/openai/test_realtime_validation.py → tests/entrypoints/openai/realtime/test_realtime_validation.py

@@ -11,11 +11,10 @@ import pybase64 as base64
 import pytest
 import websockets
 
+from tests.entrypoints.openai.conftest import add_attention_backend
+from tests.utils import ROCM_ENV_OVERRIDES, ROCM_EXTRA_ARGS, RemoteOpenAIServer
 from vllm.assets.audio import AudioAsset
 
-from ...utils import ROCM_ENV_OVERRIDES, ROCM_EXTRA_ARGS, RemoteOpenAIServer
-from .conftest import add_attention_backend
-
 MISTRAL_FORMAT_ARGS = [
     "--tokenizer_mode",
     "mistral",

tests/entrypoints/openai/responses/test_mcp_tools.py

@@ -42,7 +42,7 @@ class TestMCPToolServerUnit:
     Note: The wildcard "*" is normalized to None by
     _extract_allowed_tools_from_mcp_requests before reaching this layer,
     so we only test None and specific tool filtering here.
-    See test_serving_responses.py for "*" normalization tests.
+    See responses/test_serving_responses.py for "*" normalization tests.
     """
 
     def test_get_tool_description(self):

tests/entrypoints/openai/test_transcription_validation.py → tests/entrypoints/openai/speech_to_text/test_transcription_validation.py

@@ -6,8 +6,8 @@ import json
 
 import pytest
 
-from ...utils import ROCM_ENV_OVERRIDES, ROCM_EXTRA_ARGS, RemoteOpenAIServer
-from .conftest import add_attention_backend
+from tests.entrypoints.openai.conftest import add_attention_backend
+from tests.utils import ROCM_ENV_OVERRIDES, ROCM_EXTRA_ARGS, RemoteOpenAIServer
 
 MISTRAL_FORMAT_ARGS = [
     "--tokenizer_mode",

tests/entrypoints/openai/test_transcription_validation_whisper.py → tests/entrypoints/openai/speech_to_text/test_transcription_validation_whisper.py

@@ -13,7 +13,7 @@ import pytest
 import pytest_asyncio
 import soundfile as sf
 
-from ...utils import RemoteOpenAIServer
+from tests.utils import RemoteOpenAIServer
 
 MODEL_NAME = "openai/whisper-large-v3-turbo"
 

tests/entrypoints/openai/test_translation_validation.py → tests/entrypoints/openai/speech_to_text/test_translation_validation.py

@@ -14,8 +14,8 @@ import pytest
 import pytest_asyncio
 import soundfile as sf
 
-from ...utils import RemoteOpenAIServer
-from .conftest import add_attention_backend
+from tests.entrypoints.openai.conftest import add_attention_backend
+from tests.utils import RemoteOpenAIServer
 
 SERVER_ARGS = ["--enforce-eager"]
 

tests/entrypoints/openai/test_tokenization.py → tests/entrypoints/serve/tokenize/test_tokenization.py

@@ -5,10 +5,9 @@ import pytest
 import pytest_asyncio
 import requests
 
+from tests.utils import RemoteOpenAIServer
 from vllm.tokenizers import get_tokenizer
 
-from ...utils import RemoteOpenAIServer
-
 # any model with a chat template should work here
 MODEL_NAME = "HuggingFaceH4/zephyr-7b-beta"
 

tests/entrypoints/openai/test_tokenization_vlm.py → tests/entrypoints/serve/tokenize/test_tokenization_vlm.py

@@ -13,7 +13,7 @@ import json
 import pytest
 import requests
 
-from ...utils import RemoteOpenAIServer
+from tests.utils import RemoteOpenAIServer
 
 MODEL_NAME = "Qwen/Qwen2.5-VL-3B-Instruct"