[Frontend] Move engine/protocol.py out openai folder (#54492)

Signed-off-by: wang.yuqi <yuqi.wang@daocloud.io>
Signed-off-by: wang.yuqi <noooop@126.com>
Co-authored-by: Cyrus Leung <cyrus.tl.leung@gmail.com>

.buildkite/test_areas/misc.yaml

@@ -342,7 +342,7 @@ steps:
   - vllm/entrypoints/chat_utils.py
   - vllm/entrypoints/mcp/
   - vllm/entrypoints/openai/chat_completion/protocol.py
-  - vllm/entrypoints/openai/engine/protocol.py
+  - vllm/entrypoints/generate/base/protocol.py
   - vllm/envs.py
   - vllm/exceptions.py
   - vllm/inputs/

rust/src/server/src/routes/openai/utils/structured_outputs.rs

@@ -29,7 +29,7 @@ pub struct JsonSchemaFormat {
 /// that we can support the vLLM-specific `structural_tag` variant.
 ///
 /// Original Python definitions:
-/// <https://github.com/vllm-project/vllm/blob/f22d6e026/vllm/entrypoints/openai/engine/protocol.py#L116-L157>
+/// <https://github.com/vllm-project/vllm/blob/main/vllm/entrypoints/generate/base/protocol.py#L71-L112>
 #[derive(Debug, Clone, PartialEq, Serialize, Deserialize)]
 #[serde(tag = "type", rename_all = "snake_case")]
 pub enum ResponseFormat {

tests/entrypoints/anthropic/test_anthropic_messages_conversion.py

@@ -32,6 +32,11 @@ from vllm.entrypoints.anthropic.serving import (
     AnthropicServingMessages,
     _build_anthropic_usage,
 )
+from vllm.entrypoints.generate.base.protocol import (
+    DeltaFunctionCall,
+    DeltaMessage,
+    DeltaToolCall,
+)
 from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatCompletionResponse,
     ChatCompletionResponseChoice,
@@ -39,13 +44,7 @@ from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatCompletionStreamResponse,
     ChatMessage,
 )
-from vllm.entrypoints.openai.engine.protocol import (
-    DeltaFunctionCall,
-    DeltaMessage,
-    DeltaToolCall,
-    PromptTokenUsageInfo,
-    UsageInfo,
-)
+from vllm.entrypoints.serve.engine.protocol import PromptTokenUsageInfo, UsageInfo
 from vllm.entrypoints.serve.exception_handling.handlers.validation import (
     validation_exception_handler,
 )

tests/entrypoints/cohere/test_api_router.py

@@ -33,8 +33,8 @@ from vllm.entrypoints.cohere.protocol import (
     CohereChatV2Response,
 )
 from vllm.entrypoints.cohere.serving import CohereServingChatV2
-from vllm.entrypoints.openai.engine.protocol import ErrorInfo, ErrorResponse
 from vllm.entrypoints.scale_out.token_in_token_out.protocol import GenerateRequest
+from vllm.entrypoints.serve.engine.protocol import ErrorInfo, ErrorResponse
 from vllm.entrypoints.serve.exception_handling.handlers.http import (
     http_exception_handler,
 )

tests/entrypoints/generate/generative_scoring/test_generative_scoring.py

@@ -24,9 +24,9 @@ from vllm.entrypoints.generate.generative_scoring.serving import (
     GenerativeScoringResponse,
     ServingGenerativeScoring,
 )
-from vllm.entrypoints.openai.engine.protocol import ErrorResponse
 from vllm.entrypoints.openai.models.protocol import BaseModelPath
 from vllm.entrypoints.openai.models.serving import OpenAIServingModels
+from vllm.entrypoints.serve.engine.protocol import ErrorResponse
 from vllm.logprobs import Logprob
 from vllm.outputs import CompletionOutput, RequestOutput
 from vllm.tokenizers import get_tokenizer

tests/entrypoints/launchers/test_run_batch.py

@@ -21,7 +21,7 @@ from vllm.entrypoints.launchers.run_batch import (
     make_transcription_wrapper,
     upload_data,
 )
-from vllm.entrypoints.openai.engine.protocol import ErrorResponse
+from vllm.entrypoints.serve.engine.protocol import ErrorResponse
 from vllm.exceptions import VLLMValidationError
 from vllm.utils.mem_constants import MiB_bytes
 

tests/entrypoints/openai/chat_completion/test_batched_chat_completions.py

@@ -8,13 +8,13 @@ import httpx
 import pytest
 
 from tests.utils import RemoteOpenAIServer
+from vllm.entrypoints.generate.base.protocol import RequestResponseMetadata
 from vllm.entrypoints.openai.chat_completion.batch_serving import (
     OpenAIServingChatBatch,
 )
 from vllm.entrypoints.openai.chat_completion.protocol import (
     BatchChatCompletionRequest,
 )
-from vllm.entrypoints.openai.engine.protocol import RequestResponseMetadata
 from vllm.outputs import CompletionOutput, RequestOutput
 
 # any model with a chat template defined in tokenizer_config should work here

tests/entrypoints/openai/chat_completion/test_chat_error.py

@@ -13,11 +13,10 @@ from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatCompletionRequest,
 )
 from vllm.entrypoints.openai.chat_completion.serving import OpenAIServingChat
-from vllm.entrypoints.openai.engine.protocol import GenerationError
 from vllm.entrypoints.openai.models.protocol import BaseModelPath
 from vllm.entrypoints.openai.models.serving import OpenAIServingModels
 from vllm.entrypoints.scale_out.render.serving import ServingRender
-from vllm.exceptions import VLLMValidationError
+from vllm.exceptions import GenerationError, VLLMValidationError
 from vllm.outputs import CompletionOutput, RequestOutput
 from vllm.renderers.hf import HfRenderer
 from vllm.renderers.online_renderer import OnlineRenderer

tests/entrypoints/openai/chat_completion/test_serving_chat.py

@@ -20,6 +20,10 @@ from tests.entrypoints.openai.utils import (
 from tests.utils import RemoteOpenAIServer
 from vllm._aiter_ops import is_aiter_found_and_supported
 from vllm.config import MultiModalConfig
+from vllm.entrypoints.generate.base.protocol import (
+    DeltaMessage,
+    RequestResponseMetadata,
+)
 from vllm.entrypoints.generate.base.serving import build_per_request_timing_metrics
 from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatCompletionRequest,
@@ -31,17 +35,13 @@ from vllm.entrypoints.openai.chat_completion.serving import (
     _make_completion_tokens_details,
     _make_prompt_tokens_details,
 )
-from vllm.entrypoints.openai.engine.protocol import (
-    DeltaMessage,
-    ErrorResponse,
-    RequestResponseMetadata,
-)
 from vllm.entrypoints.openai.models.serving import (
     BaseModelPath,
     OpenAIModelRegistry,
     OpenAIServingModels,
 )
 from vllm.entrypoints.openai.parser.harmony_utils import get_encoding
+from vllm.entrypoints.serve.engine.protocol import ErrorResponse
 from vllm.exceptions import VLLMValidationError
 from vllm.inputs import TokensPrompt
 from vllm.multimodal.inputs import PlaceholderRange

tests/entrypoints/openai/completion/test_completion_error.py

@@ -8,16 +8,13 @@ from unittest.mock import AsyncMock, MagicMock
 import pytest
 
 from vllm.config.multimodal import MultiModalConfig
+from vllm.entrypoints.generate.base.protocol import RequestResponseMetadata
 from vllm.entrypoints.openai.completion.protocol import CompletionRequest
 from vllm.entrypoints.openai.completion.serving import OpenAIServingCompletion
-from vllm.entrypoints.openai.engine.protocol import (
-    GenerationError,
-    RequestResponseMetadata,
-)
 from vllm.entrypoints.openai.models.protocol import BaseModelPath
 from vllm.entrypoints.openai.models.serving import OpenAIServingModels
 from vllm.entrypoints.scale_out.render.serving import ServingRender
-from vllm.exceptions import VLLMValidationError
+from vllm.exceptions import GenerationError, VLLMValidationError
 from vllm.outputs import CompletionOutput, RequestOutput
 from vllm.renderers.hf import HfRenderer
 from vllm.renderers.online_renderer import OnlineRenderer

tests/entrypoints/openai/completion/test_lora_resolvers.py

@@ -11,9 +11,9 @@ import pytest
 from vllm.config.multimodal import MultiModalConfig
 from vllm.entrypoints.openai.completion.protocol import CompletionRequest
 from vllm.entrypoints.openai.completion.serving import OpenAIServingCompletion
-from vllm.entrypoints.openai.engine.protocol import ErrorResponse
 from vllm.entrypoints.openai.models.protocol import BaseModelPath
 from vllm.entrypoints.openai.models.serving import OpenAIServingModels
+from vllm.entrypoints.serve.engine.protocol import ErrorResponse
 from vllm.lora.request import LoRARequest
 from vllm.lora.resolver import LoRAResolver, LoRAResolverRegistry
 from vllm.renderers.hf import HfRenderer

tests/entrypoints/openai/responses/test_errors.py

@@ -7,8 +7,9 @@ from unittest.mock import MagicMock
 import pytest
 
 import vllm.envs as envs
-from vllm.entrypoints.generate.base.serving import GenerateBaseServing, GenerationError
+from vllm.entrypoints.generate.base.serving import GenerateBaseServing
 from vllm.envs import disable_envs_cache
+from vllm.exceptions import GenerationError
 
 
 @pytest.mark.asyncio

tests/entrypoints/openai/responses/test_parsable_context_unit.py

@@ -11,7 +11,7 @@ from unittest.mock import MagicMock
 
 import pytest
 
-from vllm.entrypoints.openai.engine.protocol import (
+from vllm.entrypoints.generate.base.protocol import (
     DeltaMessage,
     ExtractedToolCallInformation,
     FunctionCall,

tests/entrypoints/openai/responses/test_serving_responses.py

@@ -25,14 +25,13 @@ from openai.types.responses.tool import (
 )
 
 import vllm.envs as envs
-from vllm.entrypoints.mcp.tool_server import ToolServer
-from vllm.entrypoints.openai.engine.protocol import (
+from vllm.entrypoints.generate.base.protocol import (
     DeltaFunctionCall,
     DeltaMessage,
     DeltaToolCall,
-    ErrorResponse,
     RequestResponseMetadata,
 )
+from vllm.entrypoints.mcp.tool_server import ToolServer
 from vllm.entrypoints.openai.responses.context import ConversationContext, SimpleContext
 from vllm.entrypoints.openai.responses.protocol import (
     ResponseCreatedEvent,
@@ -49,6 +48,7 @@ from vllm.entrypoints.openai.responses.serving import (
 from vllm.entrypoints.openai.responses.streaming_events import (
     StreamingState,
 )
+from vllm.entrypoints.serve.engine.protocol import ErrorResponse
 from vllm.inputs import tokens_input
 from vllm.outputs import CompletionOutput, RequestOutput
 from vllm.parser.harmony import Segment

tests/entrypoints/openai/responses/test_streaming_events.py

@@ -1,7 +1,7 @@
 # SPDX-License-Identifier: Apache-2.0
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
 
-from vllm.entrypoints.openai.engine.protocol import (
+from vllm.entrypoints.generate.base.protocol import (
     DeltaFunctionCall,
     DeltaMessage,
     DeltaToolCall,

tests/entrypoints/openai/test_render_parity.py

@@ -26,9 +26,9 @@ from openai.types.shared import Reasoning
 
 from vllm.config.multimodal import MultiModalConfig
 from vllm.entrypoints.openai.chat_completion.protocol import ChatCompletionRequest
-from vllm.entrypoints.openai.engine.protocol import ErrorResponse
 from vllm.entrypoints.openai.responses.protocol import ResponsesRequest
 from vllm.entrypoints.openai.responses.serving import OpenAIServingResponses
+from vllm.entrypoints.serve.engine.protocol import ErrorResponse
 from vllm.inputs import tokens_input
 from vllm.renderers.online_renderer import OnlineRenderer
 from vllm.renderers.params import ChatParams

tests/entrypoints/openai/test_tool_choice_content_none.py

@@ -5,6 +5,13 @@ import pytest
 from openai.types.chat.chat_completion import ChatCompletion as OpenAIChatCompletion
 from openai.types.chat.chat_completion_chunk import ChatCompletionChunk
 
+from vllm.entrypoints.generate.base.protocol import (
+    DeltaFunctionCall,
+    DeltaMessage,
+    DeltaToolCall,
+    FunctionCall,
+    ToolCall,
+)
 from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatCompletionRequest,
     ChatCompletionResponse,
@@ -13,15 +20,8 @@ from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatCompletionStreamResponse,
     ChatMessage,
 )
-from vllm.entrypoints.openai.engine.protocol import (
-    DeltaFunctionCall,
-    DeltaMessage,
-    DeltaToolCall,
-    FunctionCall,
-    ToolCall,
-    UsageInfo,
-)
 from vllm.entrypoints.openai.responses.protocol import ResponsesRequest
+from vllm.entrypoints.serve.engine.protocol import UsageInfo
 from vllm.parser.abstract_parser import DelegatingParser
 
 pytestmark = pytest.mark.skip_global_cleanup

tests/entrypoints/openai/utils.py

@@ -10,7 +10,7 @@ from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatCompletionStreamResponse,
     ChatMessage,
 )
-from vllm.entrypoints.openai.engine.protocol import UsageInfo
+from vllm.entrypoints.serve.engine.protocol import UsageInfo
 
 
 async def accumulate_streaming_response(

tests/entrypoints/scale_out/derender/test_derender_stream.py

@@ -23,6 +23,7 @@ from vllm.entrypoints.scale_out.token_in_token_out.protocol import (
     GenerateResponseStreamChoice,
     GenerateStreamResponse,
 )
+from vllm.entrypoints.serve.engine.protocol import ErrorResponse, UsageInfo
 
 MODEL_NAME = "hmellor/tiny-random-LlamaForCausalLM"
 
@@ -40,7 +41,6 @@ def _make_stream_chunk(
     usage: dict | None = None,
 ) -> GenerateStreamResponse:
     """Build a GenerateStreamResponse SSE chunk."""
-    from vllm.entrypoints.openai.engine.protocol import UsageInfo
 
     return GenerateStreamResponse(
         request_id=request_id,
@@ -61,7 +61,6 @@ def _make_usage_chunk(
     request_id: str = "test-req",
 ) -> GenerateStreamResponse:
     """Build a usage only final SSE chunk (empty choices)."""
-    from vllm.entrypoints.openai.engine.protocol import UsageInfo
 
     return GenerateStreamResponse(
         request_id=request_id,
@@ -473,7 +472,6 @@ class TestServingDerenderStreamErrorHandling:
     @pytest.mark.asyncio
     @pytest.mark.parametrize("exc", [KeyError("bad byte"), IndexError("oob")])
     async def test_completion_stream_bad_state_returns_400(self, exc):
-        from vllm.entrypoints.openai.engine.protocol import ErrorResponse
         from vllm.entrypoints.scale_out.token_in_token_out.protocol import (
             DerenderCompletionStreamRequest,
         )
@@ -492,7 +490,6 @@ class TestServingDerenderStreamErrorHandling:
     @pytest.mark.asyncio
     @pytest.mark.parametrize("exc", [KeyError("bad byte"), IndexError("oob")])
     async def test_chat_stream_bad_state_returns_400(self, exc):
-        from vllm.entrypoints.openai.engine.protocol import ErrorResponse
         from vllm.entrypoints.scale_out.token_in_token_out.protocol import (
             DerenderChatStreamRequest,
         )

tests/entrypoints/scale_out/token_in_token_out/test_generate_stream.py

@@ -9,7 +9,7 @@ from unittest.mock import AsyncMock, MagicMock
 import pytest
 
 from vllm.config.multimodal import MultiModalConfig
-from vllm.entrypoints.openai.engine.protocol import GenerationError, StreamOptions
+from vllm.entrypoints.generate.base.protocol import StreamOptions
 from vllm.entrypoints.openai.models.protocol import BaseModelPath
 from vllm.entrypoints.openai.models.serving import OpenAIServingModels
 from vllm.entrypoints.scale_out.token_in_token_out.protocol import (
@@ -17,6 +17,7 @@ from vllm.entrypoints.scale_out.token_in_token_out.protocol import (
     GenerateResponse,
 )
 from vllm.entrypoints.scale_out.token_in_token_out.serving import ServingTokens
+from vllm.exceptions import GenerationError
 from vllm.logprobs import Logprob
 from vllm.outputs import CompletionOutput, RequestOutput
 from vllm.renderers import renderer_from_config

tests/entrypoints/serve/lora/test_serving_models.py

@@ -9,13 +9,11 @@ import pytest
 from vllm import PoolingParams
 from vllm.config import ModelConfig
 from vllm.engine.protocol import EngineClient
-from vllm.entrypoints.openai.engine.protocol import (
-    ErrorResponse,
-)
 from vllm.entrypoints.openai.models.protocol import BaseModelPath
 from vllm.entrypoints.openai.models.serving import OpenAIServingModels
 from vllm.entrypoints.pooling.base.serving import PoolingBaseServing
 from vllm.entrypoints.pooling.typing import PoolingServeContext
+from vllm.entrypoints.serve.engine.protocol import ErrorResponse
 from vllm.entrypoints.serve.lora.protocol import (
     LoadLoRAAdapterRequest,
     UnloadLoRAAdapterRequest,

tests/entrypoints/serve/utils/test_api_utils.py

@@ -5,7 +5,7 @@ from argparse import Namespace
 
 import pytest
 
-from vllm.entrypoints.openai.engine.protocol import StreamOptions
+from vllm.entrypoints.generate.base.protocol import StreamOptions
 from vllm.entrypoints.serve.utils import api_utils
 from vllm.entrypoints.serve.utils.api_utils import (
     get_max_tokens,

tests/entrypoints/speech_to_text/transcription/test_transcription_inter_chunk_spacing.py

@@ -19,11 +19,9 @@ import pytest
 
 from vllm.config import ModelConfig
 from vllm.config.speech_to_text import SpeechToTextConfig
-from vllm.entrypoints.openai.engine.protocol import (
-    ErrorResponse,
-    RequestResponseMetadata,
-)
+from vllm.entrypoints.generate.base.protocol import RequestResponseMetadata
 from vllm.entrypoints.openai.models.serving import OpenAIServingModels
+from vllm.entrypoints.serve.engine.protocol import ErrorResponse
 from vllm.entrypoints.speech_to_text.base.serving import (
     SpeechToTextBaseServing,
     asr_inter_chunk_separator,

tests/parser/engine/replay_harness.py

@@ -13,10 +13,10 @@ import json
 from collections.abc import Sequence
 from dataclasses import dataclass, field
 
+from vllm.entrypoints.generate.base.protocol import DeltaMessage
 from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatCompletionRequest,
 )
-from vllm.entrypoints.openai.engine.protocol import DeltaMessage
 
 
 @dataclass

tests/parser/engine/streaming_helpers.py

@@ -6,7 +6,7 @@ from __future__ import annotations
 
 from typing import Any
 
-from vllm.entrypoints.openai.engine.protocol import DeltaMessage
+from vllm.entrypoints.generate.base.protocol import DeltaMessage
 
 
 def _build_token_id_map(parser) -> dict[str, int]:

tests/parser/engine/test_gemma4_streaming_reasoning.py

@@ -14,10 +14,10 @@ from tests.parser.engine.streaming_helpers import (
     collect_tool_arguments,
     simulate_tool_streaming,
 )
+from vllm.entrypoints.generate.base.protocol import DeltaMessage
 from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatCompletionRequest,
 )
-from vllm.entrypoints.openai.engine.protocol import DeltaMessage
 from vllm.parser.gemma4 import Gemma4Parser
 
 # ── Special token IDs (arbitrary but consistent) ─────────────────────

tests/parser/engine/test_parser_engine.py

@@ -15,15 +15,15 @@ import pytest
 import regex as re
 
 from tests.parser.engine.conftest import make_mock_tokenizer
-from vllm.entrypoints.openai.chat_completion.protocol import (
-    ChatCompletionRequest,
-    ChatCompletionToolsParam,
-)
-from vllm.entrypoints.openai.engine.protocol import (
+from vllm.entrypoints.generate.base.protocol import (
     DeltaFunctionCall,
     DeltaToolCall,
     FunctionDefinition,
 )
+from vllm.entrypoints.openai.chat_completion.protocol import (
+    ChatCompletionRequest,
+    ChatCompletionToolsParam,
+)
 from vllm.parser.abstract_parser import DelegatingParser
 from vllm.parser.engine.adapters import make_adapters
 from vllm.parser.engine.events import EventType, SemanticEvent

tests/parser/mistral/test_tool_calls.py

@@ -25,16 +25,16 @@ from mistral_common.protocol.instruct.tool_calls import (
 )
 from partial_json_parser.core.options import Allow
 
-from vllm.entrypoints.openai.chat_completion.protocol import (
-    ChatCompletionRequest,
-)
-from vllm.entrypoints.openai.engine.protocol import (
+from vllm.entrypoints.generate.base.protocol import (
     DeltaFunctionCall,
     DeltaMessage,
     DeltaToolCall,
     ExtractedToolCallInformation,
     StructuralTagResponseFormat,
 )
+from vllm.entrypoints.openai.chat_completion.protocol import (
+    ChatCompletionRequest,
+)
 from vllm.parser.engine.events import EventType
 from vllm.parser.engine.streaming_parser_engine import StreamingParserEngine
 from vllm.parser.mistral import _DEFAULT_JSON_SCHEMA, MistralParser, mistral_config

tests/parser/test_harmony.py

@@ -15,8 +15,8 @@ from openai_harmony import (
 from transformers import AutoTokenizer, GenerationConfig
 
 from vllm.config import StructuredOutputsConfig, VllmConfig
+from vllm.entrypoints.generate.base.protocol import FunctionCall
 from vllm.entrypoints.openai.chat_completion.protocol import ChatCompletionRequest
-from vllm.entrypoints.openai.engine.protocol import FunctionCall
 from vllm.entrypoints.openai.parser.harmony_utils import (
     get_encoding,
 )

tests/parser/test_include_reasoning.py

@@ -10,8 +10,8 @@ import json
 
 import pytest
 
+from vllm.entrypoints.generate.base.protocol import DeltaMessage
 from vllm.entrypoints.openai.chat_completion.protocol import ChatCompletionRequest
-from vllm.entrypoints.openai.engine.protocol import DeltaMessage
 from vllm.entrypoints.openai.responses.protocol import ResponsesRequest
 from vllm.parser.abstract_parser import DelegatingParser
 from vllm.reasoning.basic_parsers import BaseThinkingReasoningParser

tests/parser/test_streaming.py

@@ -6,8 +6,8 @@ from types import SimpleNamespace
 
 import pytest
 
+from vllm.entrypoints.generate.base.protocol import DeltaMessage
 from vllm.entrypoints.openai.chat_completion.protocol import ChatCompletionRequest
-from vllm.entrypoints.openai.engine.protocol import DeltaMessage
 from vllm.parser.abstract_parser import DelegatingParser
 from vllm.parser.engine.registered_adapters import Qwen3ParserReasoningAdapter
 from vllm.reasoning.basic_parsers import BaseThinkingReasoningParser

tests/plugins/bge_m3_sparse_plugin/bge_m3_sparse_processor/sparse_embeddings_processor.py

@@ -4,7 +4,7 @@
 from collections.abc import Sequence
 
 from vllm.config import PoolerConfig, VllmConfig
-from vllm.entrypoints.openai.engine.protocol import UsageInfo
+from vllm.entrypoints.serve.engine.protocol import UsageInfo
 from vllm.inputs import PromptType
 from vllm.outputs import PoolingRequestOutput
 from vllm.plugins.io_processors.interface import IOProcessor

tests/plugins/bge_m3_sparse_plugin/bge_m3_sparse_processor/types.py

@@ -5,11 +5,11 @@ from typing import Literal, get_args
 
 from pydantic import BaseModel, Field
 
-from vllm.entrypoints.openai.engine.protocol import UsageInfo
 from vllm.entrypoints.pooling.base.protocol import (
     CompletionRequestMixin,
     EmbedRequestMixin,
 )
+from vllm.entrypoints.serve.engine.protocol import UsageInfo
 
 EmbedTask = Literal[
     "sparse",

tests/plugins/colbert_query_plugin/colbert_query_processor/query_embedding_processor.py

@@ -4,7 +4,7 @@ from collections.abc import Iterator, Sequence
 from typing import cast
 
 from vllm.config import VllmConfig
-from vllm.entrypoints.openai.engine.protocol import UsageInfo
+from vllm.entrypoints.serve.engine.protocol import UsageInfo
 from vllm.inputs import PromptType, TokensPrompt
 from vllm.outputs import PoolingRequestOutput
 from vllm.plugins.io_processors.interface import IOProcessor

tests/plugins/colbert_query_plugin/colbert_query_processor/types.py

@@ -5,8 +5,8 @@ from typing import Literal, get_args
 
 from pydantic import BaseModel, Field
 
-from vllm.entrypoints.openai.engine.protocol import UsageInfo
 from vllm.entrypoints.pooling.base.protocol import CompletionRequestMixin
+from vllm.entrypoints.serve.engine.protocol import UsageInfo
 
 InputType = Literal["query", "document"]
 INPUT_TYPES: tuple[InputType, ...] = get_args(InputType)

tests/reasoning/test_cohere_command_reasoning_parser.py

@@ -12,14 +12,14 @@ from typing import Any
 import pytest
 from pydantic import ValidationError
 
-from vllm.entrypoints.openai.chat_completion.protocol import (
-    ChatCompletionRequest,
-)
-from vllm.entrypoints.openai.engine.protocol import (
+from vllm.entrypoints.generate.base.protocol import (
     JsonSchemaResponseFormat,
     ResponseFormat,
     StructuralTagResponseFormat,
 )
+from vllm.entrypoints.openai.chat_completion.protocol import (
+    ChatCompletionRequest,
+)
 from vllm.entrypoints.openai.responses.protocol import ResponsesRequest
 from vllm.reasoning.cohere_command_reasoning_parser import (
     CohereCommand3ReasoningParser,

tests/reasoning/test_deepseekv3_reasoning_parser.py

@@ -4,8 +4,8 @@
 import pytest
 from transformers import AutoTokenizer
 
+from vllm.entrypoints.generate.base.protocol import DeltaMessage
 from vllm.entrypoints.openai.chat_completion.protocol import ChatCompletionRequest
-from vllm.entrypoints.openai.engine.protocol import DeltaMessage
 from vllm.reasoning import ReasoningParserManager
 from vllm.reasoning.deepseek_r1_reasoning_parser import DeepSeekR1ReasoningParser
 from vllm.reasoning.deepseek_v3_reasoning_parser import DeepSeekV3ReasoningParser

tests/reasoning/test_kimi_k2_reasoning_parser.py

@@ -5,8 +5,8 @@ from unittest.mock import MagicMock
 
 import pytest
 
+from vllm.entrypoints.generate.base.protocol import DeltaMessage
 from vllm.entrypoints.openai.chat_completion.protocol import ChatCompletionRequest
-from vllm.entrypoints.openai.engine.protocol import DeltaMessage
 from vllm.reasoning.kimi_k2_reasoning_parser import KimiK2ReasoningParser
 from vllm.tokenizers import get_tokenizer
 

tests/reasoning/test_kimi_k3_reasoning_parser.py

@@ -3,10 +3,10 @@
 
 import pytest
 
+from vllm.entrypoints.generate.base.protocol import DeltaMessage
 from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatCompletionRequest,
 )
-from vllm.entrypoints.openai.engine.protocol import DeltaMessage
 from vllm.parser.kimi_k3 import KimiK3Parser
 from vllm.parser.parser_manager import ParserManager
 from vllm.reasoning.kimi_k3_reasoning_parser import KimiK3ReasoningParser

tests/reasoning/utils.py

@@ -1,8 +1,8 @@
 # SPDX-License-Identifier: Apache-2.0
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
 
+from vllm.entrypoints.generate.base.protocol import DeltaMessage
 from vllm.entrypoints.openai.chat_completion.protocol import ChatCompletionRequest
-from vllm.entrypoints.openai.engine.protocol import DeltaMessage
 from vllm.reasoning import ReasoningParser
 from vllm.utils.mistral import is_mistral_tokenizer
 

tests/tool_parsers/test_dots_tool_parser.py

@@ -7,11 +7,11 @@ from unittest.mock import MagicMock
 
 import pytest
 
+from vllm.entrypoints.generate.base.protocol import DeltaMessage
 from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatCompletionRequest,
     ChatCompletionToolsParam,
 )
-from vllm.entrypoints.openai.engine.protocol import DeltaMessage
 from vllm.tool_parsers import ToolParserManager
 from vllm.tool_parsers.dots_tool_parser import DotsToolParser
 

tests/tool_parsers/test_ernie45_moe_tool_parser.py

@@ -7,12 +7,12 @@ from collections.abc import Generator
 
 import pytest
 
-from vllm.entrypoints.openai.chat_completion.protocol import ChatCompletionRequest
-from vllm.entrypoints.openai.engine.protocol import (
+from vllm.entrypoints.generate.base.protocol import (
     DeltaMessage,
     FunctionCall,
     ToolCall,
 )
+from vllm.entrypoints.openai.chat_completion.protocol import ChatCompletionRequest
 from vllm.tokenizers import TokenizerLike, get_tokenizer
 from vllm.tokenizers.detokenizer_utils import detokenize_incrementally
 from vllm.tool_parsers.ernie45_tool_parser import Ernie45ToolParser

tests/tool_parsers/test_gigachat3_tool_parser.py

@@ -10,7 +10,7 @@ from tests.tool_parsers.utils import (
     run_tool_extraction,
     run_tool_extraction_streaming,
 )
-from vllm.entrypoints.openai.engine.protocol import FunctionCall
+from vllm.entrypoints.generate.base.protocol import FunctionCall
 from vllm.tokenizers import TokenizerLike
 from vllm.tool_parsers import ToolParser, ToolParserManager
 

tests/tool_parsers/test_glm47_moe_tool_parser.py

@@ -9,12 +9,12 @@ from unittest.mock import Mock
 import pytest
 from openai.types.responses import ResponseFunctionToolCall
 
+from vllm.entrypoints.generate.base.protocol import FunctionCall
 from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatCompletionRequest,
     ChatCompletionToolsParam,
     FunctionDefinition,
 )
-from vllm.entrypoints.openai.engine.protocol import FunctionCall
 from vllm.entrypoints.openai.responses.protocol import ResponsesRequest
 from vllm.entrypoints.openai.responses.utils import build_response_output_items
 from vllm.tokenizers import get_tokenizer

tests/tool_parsers/test_granite4_tool_parser.py

@@ -7,10 +7,10 @@ from typing import Any
 import pytest
 from transformers import AutoTokenizer
 
-from vllm.entrypoints.openai.chat_completion.protocol import ChatCompletionRequest
-from vllm.entrypoints.openai.engine.protocol import (
+from vllm.entrypoints.generate.base.protocol import (
     DeltaMessage,
 )
+from vllm.entrypoints.openai.chat_completion.protocol import ChatCompletionRequest
 from vllm.tool_parsers.granite4_tool_parser import Granite4ToolParser
 
 MODEL = "ibm-granite/granite-4.0-h-tiny"

tests/tool_parsers/test_hunyuan_a13b_tool_parser.py

@@ -11,7 +11,7 @@ from tests.tool_parsers.utils import (
     run_tool_extraction,
     run_tool_extraction_streaming,
 )
-from vllm.entrypoints.openai.engine.protocol import FunctionCall, ToolCall
+from vllm.entrypoints.generate.base.protocol import FunctionCall, ToolCall
 from vllm.tool_parsers import ToolParser, ToolParserManager
 
 

tests/tool_parsers/test_hy_v3_tool_parser.py

@@ -8,12 +8,12 @@ from unittest.mock import Mock
 
 import pytest
 
+from vllm.entrypoints.generate.base.protocol import DeltaMessage
 from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatCompletionRequest,
     ChatCompletionToolsParam,
     FunctionDefinition,
 )
-from vllm.entrypoints.openai.engine.protocol import DeltaMessage
 from vllm.tokenizers import get_tokenizer
 from vllm.tool_parsers.hy_v3_tool_parser import HYV3ToolParser
 

tests/tool_parsers/test_jamba_tool_parser.py

@@ -8,7 +8,7 @@ import partial_json_parser
 import pytest
 from partial_json_parser.core.options import Allow
 
-from vllm.entrypoints.openai.engine.protocol import DeltaMessage, FunctionCall, ToolCall
+from vllm.entrypoints.generate.base.protocol import DeltaMessage, FunctionCall, ToolCall
 from vllm.tokenizers import TokenizerLike, get_tokenizer
 from vllm.tokenizers.detokenizer_utils import detokenize_incrementally
 from vllm.tool_parsers.jamba_tool_parser import JambaToolParser

tests/tool_parsers/test_lfm2_tool_parser.py

@@ -10,7 +10,7 @@ from tests.tool_parsers.utils import (
     run_tool_extraction,
     run_tool_extraction_streaming,
 )
-from vllm.entrypoints.openai.engine.protocol import FunctionCall
+from vllm.entrypoints.generate.base.protocol import FunctionCall
 from vllm.tokenizers import TokenizerLike
 from vllm.tool_parsers import ToolParser, ToolParserManager
 

tests/tool_parsers/test_llama3_json_tool_parser.py

@@ -6,7 +6,7 @@ from unittest.mock import MagicMock, patch
 import pytest
 from transformers import AutoTokenizer
 
-from vllm.entrypoints.openai.engine.protocol import ExtractedToolCallInformation
+from vllm.entrypoints.generate.base.protocol import ExtractedToolCallInformation
 from vllm.tool_parsers.llama_tool_parser import Llama3JsonToolParser
 
 LLAMA_MODEL = "meta-llama/Llama-3.2-1B-Instruct"

tests/tool_parsers/test_llama4_pythonic_tool_parser.py

@@ -9,7 +9,7 @@ from tests.tool_parsers.utils import (
     run_tool_extraction,
     run_tool_extraction_streaming,
 )
-from vllm.entrypoints.openai.engine.protocol import FunctionCall
+from vllm.entrypoints.generate.base.protocol import FunctionCall
 from vllm.tokenizers import TokenizerLike
 from vllm.tool_parsers import ToolParser, ToolParserManager
 

tests/tool_parsers/test_minicpm5xml_tool_parser.py

@@ -9,11 +9,11 @@ from unittest.mock import MagicMock
 import pytest
 
 from tests.tool_parsers.utils import run_tool_extraction_streaming
+from vllm.entrypoints.generate.base.protocol import FunctionCall, ToolCall
 from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatCompletionRequest,
     ChatCompletionToolsParam,
 )
-from vllm.entrypoints.openai.engine.protocol import FunctionCall, ToolCall
 from vllm.tool_parsers import ToolParser, ToolParserManager
 from vllm.tool_parsers.minicpm5xml_tool_parser import MiniCPM5XMLToolParser
 

tests/tool_parsers/test_minimax_m3_tool_parser.py

@@ -6,11 +6,11 @@ from typing import Any
 
 import pytest
 
+from vllm.entrypoints.generate.base.protocol import DeltaMessage
 from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatCompletionToolsParam,
     FunctionDefinition,
 )
-from vllm.entrypoints.openai.engine.protocol import DeltaMessage
 from vllm.tool_parsers import ToolParserManager
 from vllm.tool_parsers.minimax_m3_tool_parser import MinimaxM3ToolParser
 

tests/tool_parsers/test_olmo3_tool_parser.py

@@ -9,7 +9,7 @@ from tests.tool_parsers.utils import (
     run_tool_extraction,
     run_tool_extraction_streaming,
 )
-from vllm.entrypoints.openai.engine.protocol import FunctionCall
+from vllm.entrypoints.generate.base.protocol import FunctionCall
 from vllm.tokenizers import TokenizerLike
 from vllm.tool_parsers import ToolParser, ToolParserManager
 

tests/tool_parsers/test_pythonic_tool_parser.py

@@ -9,7 +9,7 @@ from tests.tool_parsers.utils import (
     run_tool_extraction,
     run_tool_extraction_streaming,
 )
-from vllm.entrypoints.openai.engine.protocol import FunctionCall
+from vllm.entrypoints.generate.base.protocol import FunctionCall
 from vllm.tokenizers import TokenizerLike
 from vllm.tool_parsers import ToolParser, ToolParserManager
 

tests/tool_parsers/test_qwen3coder_tool_parser.py

@@ -9,6 +9,11 @@ import pytest
 from openai.types.responses.function_tool import FunctionTool
 from xgrammar import StructuralTag
 
+from vllm.entrypoints.generate.base.protocol import (
+    DeltaMessage,
+    FunctionCall,
+    ToolCall,
+)
 from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatCompletionNamedFunction,
     ChatCompletionNamedToolChoiceParam,
@@ -16,11 +21,6 @@ from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatCompletionToolsParam,
     FunctionDefinition,
 )
-from vllm.entrypoints.openai.engine.protocol import (
-    DeltaMessage,
-    FunctionCall,
-    ToolCall,
-)
 from vllm.parser.abstract_parser import DelegatingParser
 from vllm.tokenizers import TokenizerLike, get_tokenizer
 from vllm.tokenizers.detokenizer_utils import detokenize_incrementally

tests/tool_parsers/test_step3p5_tool_parser.py

@@ -6,15 +6,15 @@ from collections.abc import Generator
 
 import pytest
 
-from vllm.entrypoints.openai.chat_completion.protocol import (
-    ChatCompletionRequest,
-    ChatCompletionToolsParam,
-)
-from vllm.entrypoints.openai.engine.protocol import (
+from vllm.entrypoints.generate.base.protocol import (
     DeltaMessage,
     FunctionCall,
     ToolCall,
 )
+from vllm.entrypoints.openai.chat_completion.protocol import (
+    ChatCompletionRequest,
+    ChatCompletionToolsParam,
+)
 from vllm.tokenizers import TokenizerLike, get_tokenizer
 from vllm.tokenizers.detokenizer_utils import detokenize_incrementally
 from vllm.tool_parsers.step3p5_tool_parser import Step3p5ToolParser

tests/tool_parsers/test_xlam_tool_parser.py

@@ -6,12 +6,12 @@ from collections.abc import Generator
 
 import pytest
 
-from vllm.entrypoints.openai.chat_completion.protocol import ChatCompletionRequest
-from vllm.entrypoints.openai.engine.protocol import (
+from vllm.entrypoints.generate.base.protocol import (
     DeltaMessage,
     FunctionCall,
     ToolCall,
 )
+from vllm.entrypoints.openai.chat_completion.protocol import ChatCompletionRequest
 from vllm.tokenizers import TokenizerLike, get_tokenizer
 from vllm.tokenizers.detokenizer_utils import detokenize_incrementally
 from vllm.tool_parsers.xlam_tool_parser import xLAMToolParser

tests/tool_parsers/utils.py

@@ -3,13 +3,13 @@
 
 from collections.abc import Iterable
 
-from vllm.entrypoints.openai.chat_completion.protocol import ChatCompletionRequest
-from vllm.entrypoints.openai.engine.protocol import (
+from vllm.entrypoints.generate.base.protocol import (
     DeltaMessage,
     ExtractedToolCallInformation,
     FunctionCall,
     ToolCall,
 )
+from vllm.entrypoints.openai.chat_completion.protocol import ChatCompletionRequest
 from vllm.tokenizers import TokenizerLike
 from vllm.tool_parsers import ToolParser
 

tests/tool_use/test_kimi_k3_tool_parser.py

@@ -5,10 +5,10 @@ import json
 
 import pytest
 
+from vllm.entrypoints.generate.base.protocol import DeltaMessage
 from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatCompletionRequest,
 )
-from vllm.entrypoints.openai.engine.protocol import DeltaMessage
 from vllm.entrypoints.openai.responses.protocol import ResponsesRequest
 from vllm.entrypoints.openai.responses.utils import build_response_output_items
 from vllm.exceptions import VLLMValidationError

vllm/entrypoints/anthropic/api_router.py

@@ -16,7 +16,7 @@ from vllm.entrypoints.anthropic.protocol import (
     AnthropicMessagesResponse,
 )
 from vllm.entrypoints.anthropic.serving import AnthropicServingMessages
-from vllm.entrypoints.openai.engine.protocol import ErrorResponse
+from vllm.entrypoints.serve.engine.protocol import ErrorResponse
 from vllm.entrypoints.serve.exception_handling.error_response import (
     create_error_response,
 )

vllm/entrypoints/anthropic/serving.py

@@ -30,6 +30,11 @@ from vllm.entrypoints.anthropic.protocol import (
     AnthropicUsage,
 )
 from vllm.entrypoints.chat_utils import ChatTemplateContentFormatOption
+from vllm.entrypoints.generate.base.protocol import (
+    JsonSchemaResponseFormat,
+    ResponseFormat,
+    StreamOptions,
+)
 from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatCompletionNamedToolChoiceParam,
     ChatCompletionRequest,
@@ -38,14 +43,8 @@ from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatCompletionToolsParam,
 )
 from vllm.entrypoints.openai.chat_completion.serving import OpenAIServingChat
-from vllm.entrypoints.openai.engine.protocol import (
-    ErrorResponse,
-    JsonSchemaResponseFormat,
-    ResponseFormat,
-    StreamOptions,
-    UsageInfo,
-)
 from vllm.entrypoints.openai.models.serving import OpenAIServingModels
+from vllm.entrypoints.serve.engine.protocol import ErrorResponse, UsageInfo
 from vllm.entrypoints.serve.exception_handling.utils import sanitize_message
 from vllm.entrypoints.serve.utils.request_logger import RequestLogger
 from vllm.renderers.online_renderer import OnlineRenderer

vllm/entrypoints/cohere/api_router.py

@@ -38,7 +38,7 @@ from fastapi.responses import JSONResponse, Response, StreamingResponse
 from starlette.middleware.base import BaseHTTPMiddleware
 
 import vllm.envs as envs
-from vllm.entrypoints.openai.engine.protocol import ErrorResponse
+from vllm.entrypoints.serve.engine.protocol import ErrorResponse
 from vllm.entrypoints.serve.exception_handling.utils import sanitize_message
 from vllm.entrypoints.serve.utils.api_utils import (
     load_aware_call,

vllm/entrypoints/cohere/cohere_chat_message.py

@@ -29,11 +29,9 @@ from typing import Any, Literal
 
 from pydantic import Field, model_serializer
 
+from vllm.entrypoints.generate.base.protocol import DeltaMessage
 from vllm.entrypoints.openai.chat_completion.protocol import ChatMessage
-from vllm.entrypoints.openai.engine.protocol import (
-    DeltaMessage,
-    OpenAIBaseModel,
-)
+from vllm.entrypoints.serve.engine.protocol import OpenAIBaseModel
 
 
 class CitationSource(OpenAIBaseModel):

vllm/entrypoints/cohere/serving.py

@@ -71,6 +71,11 @@ from vllm.entrypoints.cohere.protocol import (
     ToolCallStartEvent,
     ToolPlanDeltaEvent,
 )
+from vllm.entrypoints.generate.base.protocol import (
+    JsonSchemaResponseFormat,
+    ResponseFormat,
+    StreamOptions,
+)
 from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatCompletionRequest,
     ChatCompletionResponse,
@@ -79,13 +84,8 @@ from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatMessage,
 )
 from vllm.entrypoints.openai.chat_completion.serving import OpenAIServingChat
-from vllm.entrypoints.openai.engine.protocol import (
-    ErrorResponse,
-    JsonSchemaResponseFormat,
-    ResponseFormat,
-    StreamOptions,
-)
 from vllm.entrypoints.openai.models.serving import OpenAIServingModels
+from vllm.entrypoints.serve.engine.protocol import ErrorInfo, ErrorResponse
 from vllm.entrypoints.serve.exception_handling.utils import sanitize_message
 from vllm.entrypoints.serve.utils.request_logger import RequestLogger
 from vllm.parser.abstract_parser import Parser
@@ -1761,7 +1761,6 @@ class CohereServingChatV2(OpenAIServingChat):
         # so the router can translate the envelope uniformly. ``param``
         # is accepted for signature parity with the base class but is
         # not surfaced in the Cohere wire format.
-        from vllm.entrypoints.openai.engine.protocol import ErrorInfo
 
         del param  # unused; kept for signature compatibility
         return ErrorResponse(

vllm/entrypoints/generate/base/protocol.py

@@ -4,9 +4,7 @@
 # Adapted from
 # https://github.com/lm-sys/FastChat/blob/168ccc29d3f7edc50823016105c024fe2282732a/fastchat/protocol/openai_api_protocol.py
 import json
-import time
-from http import HTTPStatus
-from typing import Annotated, Any, ClassVar, Literal, TypeAlias
+from typing import Annotated, Any, Literal, TypeAlias
 
 import regex as re
 from pydantic import (
@@ -14,16 +12,15 @@ from pydantic import (
     ConfigDict,
     Field,
     model_serializer,
-    model_validator,
 )
 
 import vllm.envs as envs
 from vllm.config.utils import replace
 from vllm.entrypoints.chat_utils import make_tool_call_id
-from vllm.exceptions import VLLMServerError, VLLMValidationError
+from vllm.entrypoints.serve.engine.protocol import OpenAIBaseModel, UsageInfo
+from vllm.exceptions import VLLMValidationError
 from vllm.logger import init_logger
 from vllm.sampling_params import StructuredOutputsParams
-from vllm.utils import random_uuid
 from vllm.utils.import_utils import resolve_obj_by_qualname
 
 logger = init_logger(__name__)
@@ -33,102 +30,6 @@ StopParam: TypeAlias = (
 )
 
 
-class OpenAIBaseModel(BaseModel):
-    # OpenAI API does allow extra fields
-    model_config = ConfigDict(extra="allow")
-
-    # Cache class field names
-    field_names: ClassVar[set[str] | None] = None
-
-    @model_validator(mode="wrap")
-    @classmethod
-    def __log_extra_fields__(cls, data, handler):
-        result = handler(data)
-        if not isinstance(data, dict):
-            return result
-        field_names = cls.field_names
-        if field_names is None:
-            # Get all class field names and their potential aliases
-            field_names = set()
-            for field_name, field in cls.model_fields.items():
-                field_names.add(field_name)
-                if alias := getattr(field, "alias", None):
-                    field_names.add(alias)
-            cls.field_names = field_names
-
-        # Compare against both field names and aliases
-        if any(k not in field_names for k in data):
-            logger.debug(
-                "The following fields were present in the request but ignored: %s",
-                data.keys() - field_names,
-            )
-        return result
-
-
-class ErrorInfo(OpenAIBaseModel):
-    message: str
-    type: str
-    param: str | None = None
-    code: int
-
-
-class ErrorResponse(OpenAIBaseModel):
-    error: ErrorInfo
-
-
-class ModelPermission(OpenAIBaseModel):
-    id: str = Field(default_factory=lambda: f"modelperm-{random_uuid()}")
-    object: str = "model_permission"
-    created: int = Field(default_factory=lambda: int(time.time()))
-    allow_create_engine: bool = False
-    allow_sampling: bool = True
-    allow_logprobs: bool = True
-    allow_search_indices: bool = False
-    allow_view: bool = True
-    allow_fine_tuning: bool = False
-    organization: str = "*"
-    group: str | None = None
-    is_blocking: bool = False
-
-
-class ModelCard(OpenAIBaseModel):
-    id: str
-    object: str = "model"
-    created: int = Field(default_factory=lambda: int(time.time()))
-    owned_by: str = "vllm"
-    root: str | None = None
-    parent: str | None = None
-    max_model_len: int | None = None
-    permission: list[ModelPermission] = Field(default_factory=list)
-
-
-class ModelList(OpenAIBaseModel):
-    object: str = "list"
-    data: list[ModelCard] = Field(default_factory=list)
-
-
-class PromptTokenUsageInfo(OpenAIBaseModel):
-    cached_tokens: int | None = None
-    created_cache_tokens: int | None = None
-    multimodal_tokens: dict[str, int] | None = None
-    """Prompt tokens contributed by each input modality, keyed by modality name
-    (e.g. `image`, `audio`, `video`). A breakdown of the multimodal
-    placeholder tokens already counted in `prompt_tokens`; `None` when the
-    request has no multimodal input."""
-
-
-class CompletionTokenUsageInfo(OpenAIBaseModel):
-    reasoning_tokens: int = 0
-
-
-class UsageInfo(OpenAIBaseModel):
-    prompt_tokens: int = 0
-    total_tokens: int = 0
-    completion_tokens: int | None = 0
-    prompt_tokens_details: PromptTokenUsageInfo | None = None
-    completion_tokens_details: CompletionTokenUsageInfo | None = None
-
-
 class SpeculativeDecodingMetrics(OpenAIBaseModel):
     """Per-request speculative-decoding acceptance metrics.
 
@@ -437,11 +338,3 @@ class DeltaMessage(OpenAIBaseModel):
         if len(data.get("tool_calls", [])) == 0:
             data.pop("tool_calls", None)
         return data
-
-
-class GenerationError(VLLMServerError):
-    """raised when finish_reason indicates internal server error (500)"""
-
-    def __init__(self, message: str = "Internal server error"):
-        super().__init__(message)
-        self.status_code = HTTPStatus.INTERNAL_SERVER_ERROR

vllm/entrypoints/generate/base/serving.py

@@ -5,27 +5,28 @@ import time
 from collections.abc import Awaitable, Mapping
 from dataclasses import dataclass, field
 from http import HTTPStatus
-from typing import TYPE_CHECKING, ClassVar, Generic, TypeVar
+from typing import ClassVar, Generic, TypeVar
 
 from fastapi import Request
 from pydantic import ConfigDict
 from starlette.datastructures import Headers
 
+from vllm import RequestOutput
 from vllm.engine.protocol import EngineClient
-from vllm.entrypoints.generate.beam_search.online import BeamSearchOnlineMixin
-from vllm.entrypoints.openai.chat_completion.protocol import ChatCompletionRequest
-from vllm.entrypoints.openai.completion.protocol import CompletionRequest
-from vllm.entrypoints.openai.engine.protocol import (
-    ErrorResponse,
-    GenerationError,
+from vllm.entrypoints.generate.base.protocol import (
     PerRequestMetrics,
     SpeculativeDecodingMetrics,
 )
+from vllm.entrypoints.generate.beam_search.online import BeamSearchOnlineMixin
+from vllm.entrypoints.openai.chat_completion.protocol import ChatCompletionRequest
+from vllm.entrypoints.openai.completion.protocol import CompletionRequest
 from vllm.entrypoints.openai.models.serving import OpenAIServingModels
 from vllm.entrypoints.openai.responses.protocol import ResponsesRequest
+from vllm.entrypoints.serve.engine.protocol import ErrorResponse
 from vllm.entrypoints.serve.engine.serving import BaseServing
 from vllm.entrypoints.serve.engine.typing import AnyRequest
 from vllm.entrypoints.serve.utils.request_logger import RequestLogger
+from vllm.exceptions import GenerationError
 from vllm.inputs import EngineInput
 from vllm.logger import init_logger
 from vllm.logprobs import Logprob, PromptLogprobs
@@ -38,9 +39,6 @@ from vllm.tracing import (
 )
 from vllm.v1.metrics.stats import RequestStateStats
 
-if TYPE_CHECKING:
-    from vllm.outputs import RequestOutput
-
 logger = init_logger(__name__)
 
 RequestT = TypeVar("RequestT", bound=AnyRequest)

vllm/entrypoints/generate/generative_scoring/api_router.py

@@ -9,7 +9,7 @@ from vllm.entrypoints.generate.generative_scoring.serving import (
     GenerativeScoringResponse,
     ServingGenerativeScoring,
 )
-from vllm.entrypoints.openai.engine.protocol import ErrorResponse
+from vllm.entrypoints.serve.engine.protocol import ErrorResponse
 from vllm.entrypoints.serve.utils.api_utils import (
     load_aware_call,
     validate_json_request,

vllm/entrypoints/generate/generative_scoring/serving.py

@@ -18,12 +18,12 @@ from fastapi import Request
 from pydantic import Field
 
 from vllm.engine.protocol import EngineClient
-from vllm.entrypoints.openai.engine.protocol import (
+from vllm.entrypoints.openai.models.serving import OpenAIServingModels
+from vllm.entrypoints.serve.engine.protocol import (
     ErrorResponse,
     OpenAIBaseModel,
     UsageInfo,
 )
-from vllm.entrypoints.openai.models.serving import OpenAIServingModels
 from vllm.entrypoints.serve.engine.serving import BaseServing
 from vllm.entrypoints.serve.utils.request_logger import RequestLogger
 from vllm.inputs import EngineInput, tokens_input

vllm/entrypoints/launchers/run_batch.py

@@ -35,11 +35,6 @@ from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatCompletionRequest,
     ChatCompletionResponse,
 )
-from vllm.entrypoints.openai.engine.protocol import (
-    ErrorInfo,
-    ErrorResponse,
-    OpenAIBaseModel,
-)
 from vllm.entrypoints.pooling.embed.protocol import (
     EmbeddingRequest,
     EmbeddingResponse,
@@ -51,6 +46,11 @@ from vllm.entrypoints.pooling.scoring.protocol import (
     ScoreResponse,
 )
 from vllm.entrypoints.serve import create_error_response
+from vllm.entrypoints.serve.engine.protocol import (
+    ErrorInfo,
+    ErrorResponse,
+    OpenAIBaseModel,
+)
 from vllm.entrypoints.speech_to_text.transcription.protocol import (
     TranscriptionRequest,
     TranscriptionResponse,

vllm/entrypoints/openai/chat_completion/api_router.py

@@ -14,8 +14,8 @@ from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatCompletionResponse,
 )
 from vllm.entrypoints.openai.chat_completion.serving import OpenAIServingChat
-from vllm.entrypoints.openai.engine.protocol import ErrorResponse
 from vllm.entrypoints.openai.sse_keep_alive import with_sse_keep_alive
+from vllm.entrypoints.serve.engine.protocol import ErrorResponse
 from vllm.entrypoints.serve.utils.api_utils import (
     load_aware_call,
     validate_json_request,

vllm/entrypoints/openai/chat_completion/batch_serving.py

@@ -9,6 +9,7 @@ from http import HTTPStatus
 from fastapi import Request
 
 from vllm.entrypoints.chat_utils import ConversationMessage
+from vllm.entrypoints.generate.base.protocol import RequestResponseMetadata
 from vllm.entrypoints.openai.chat_completion.protocol import (
     BatchChatCompletionRequest,
     ChatCompletionResponse,
@@ -16,11 +17,7 @@ from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatMessage,
 )
 from vllm.entrypoints.openai.chat_completion.serving import OpenAIServingChat
-from vllm.entrypoints.openai.engine.protocol import (
-    ErrorResponse,
-    RequestResponseMetadata,
-    UsageInfo,
-)
+from vllm.entrypoints.serve.engine.protocol import ErrorResponse, UsageInfo
 from vllm.entrypoints.serve.utils.api_utils import get_max_tokens
 from vllm.inputs import EngineInput
 from vllm.logger import init_logger

vllm/entrypoints/openai/chat_completion/protocol.py

@@ -23,21 +23,20 @@ from vllm.entrypoints.chat_utils import (
     ChatCompletionMessageParam,
     ChatTemplateContentFormatOption,
 )
-from vllm.entrypoints.openai.engine.protocol import (
+from vllm.entrypoints.generate.base.protocol import (
     AnyResponseFormat,
     DeltaMessage,
     FunctionCall,
     FunctionDefinition,
-    OpenAIBaseModel,
     PerRequestMetrics,
     StopParam,
     StreamOptions,
     ToolCall,
-    UsageInfo,
     structured_outputs_from_response_format,
     validate_structural_tag_response_format,
     validate_structured_outputs_structural_tag,
 )
+from vllm.entrypoints.serve.engine.protocol import OpenAIBaseModel, UsageInfo
 from vllm.exceptions import VLLMValidationError
 from vllm.logger import init_logger
 from vllm.logprobs import Logprob

vllm/entrypoints/openai/chat_completion/serving.py

@@ -16,9 +16,15 @@ from vllm.entrypoints.chat_utils import (
     ConversationMessage,
     make_tool_call_id,
 )
+from vllm.entrypoints.generate.base.protocol import (
+    DeltaMessage,
+    FunctionCall,
+    PerRequestMetrics,
+    RequestResponseMetadata,
+    ToolCall,
+)
 from vllm.entrypoints.generate.base.serving import (
     GenerateBaseServing,
-    GenerationError,
     build_per_request_timing_metrics,
     build_spec_decoding_metrics,
     clamp_prompt_logprobs,
@@ -36,23 +42,19 @@ from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatCompletionStreamResponse,
     ChatMessage,
 )
-from vllm.entrypoints.openai.engine.protocol import (
+from vllm.entrypoints.openai.models.serving import OpenAIServingModels
+from vllm.entrypoints.serve.engine.protocol import (
     CompletionTokenUsageInfo,
-    DeltaMessage,
     ErrorResponse,
-    FunctionCall,
-    PerRequestMetrics,
     PromptTokenUsageInfo,
-    RequestResponseMetadata,
-    ToolCall,
     UsageInfo,
 )
-from vllm.entrypoints.openai.models.serving import OpenAIServingModels
 from vllm.entrypoints.serve.utils.api_utils import get_max_tokens, should_include_usage
 from vllm.entrypoints.serve.utils.request_logger import RequestLogger
 from vllm.entrypoints.serve.utils.tool_calls_utils import (
     maybe_filter_parallel_tool_calls,
 )
+from vllm.exceptions import GenerationError
 from vllm.inputs import EngineInput, MultiModalPlaceholders
 from vllm.logger import init_logger
 from vllm.logprobs import Logprob

vllm/entrypoints/openai/completion/api_router.py

@@ -12,8 +12,8 @@ from vllm.entrypoints.openai.completion.protocol import (
     CompletionResponse,
 )
 from vllm.entrypoints.openai.completion.serving import OpenAIServingCompletion
-from vllm.entrypoints.openai.engine.protocol import ErrorResponse
 from vllm.entrypoints.openai.sse_keep_alive import with_sse_keep_alive
+from vllm.entrypoints.serve.engine.protocol import ErrorResponse
 from vllm.entrypoints.serve.utils.api_utils import (
     load_aware_call,
     validate_json_request,

vllm/entrypoints/openai/completion/protocol.py

@@ -10,17 +10,16 @@ from pydantic import Field, model_validator
 
 import vllm.envs as envs
 from vllm.config import ModelConfig
-from vllm.entrypoints.openai.engine.protocol import (
+from vllm.entrypoints.generate.base.protocol import (
     AnyResponseFormat,
-    OpenAIBaseModel,
     PerRequestMetrics,
     StopParam,
     StreamOptions,
-    UsageInfo,
     structured_outputs_from_response_format,
     validate_structural_tag_response_format,
     validate_structured_outputs_structural_tag,
 )
+from vllm.entrypoints.serve.engine.protocol import OpenAIBaseModel, UsageInfo
 from vllm.exceptions import VLLMValidationError
 from vllm.logger import init_logger
 from vllm.logprobs import Logprob

vllm/entrypoints/openai/completion/serving.py

@@ -10,9 +10,12 @@ from typing import cast
 from fastapi import Request
 
 from vllm.engine.protocol import EngineClient
+from vllm.entrypoints.generate.base.protocol import (
+    PerRequestMetrics,
+    RequestResponseMetadata,
+)
 from vllm.entrypoints.generate.base.serving import (
     GenerateBaseServing,
-    GenerationError,
     build_per_request_timing_metrics,
     build_spec_decoding_metrics,
     clamp_prompt_logprobs,
@@ -26,17 +29,15 @@ from vllm.entrypoints.openai.completion.protocol import (
     CompletionResponseStreamChoice,
     CompletionStreamResponse,
 )
-from vllm.entrypoints.openai.engine.protocol import (
+from vllm.entrypoints.openai.models.serving import OpenAIServingModels
+from vllm.entrypoints.serve.engine.protocol import (
     ErrorResponse,
-    PerRequestMetrics,
     PromptTokenUsageInfo,
-    RequestResponseMetadata,
     UsageInfo,
 )
-from vllm.entrypoints.openai.models.serving import OpenAIServingModels
 from vllm.entrypoints.serve.utils.api_utils import get_max_tokens, should_include_usage
 from vllm.entrypoints.serve.utils.request_logger import RequestLogger
-from vllm.exceptions import VLLMValidationError
+from vllm.exceptions import GenerationError, VLLMValidationError
 from vllm.inputs import EngineInput
 from vllm.logger import init_logger
 from vllm.logprobs import Logprob

vllm/entrypoints/openai/engine/__init__.py

@@ -1,2 +0,0 @@
-# SPDX-License-Identifier: Apache-2.0
-# SPDX-FileCopyrightText: Copyright contributors to the vLLM project

vllm/entrypoints/openai/models/serving.py

@@ -7,14 +7,14 @@ from http import HTTPStatus
 
 from vllm.config import ModelConfig
 from vllm.engine.protocol import EngineClient
-from vllm.entrypoints.openai.engine.protocol import (
+from vllm.entrypoints.openai.models.protocol import BaseModelPath, LoRAModulePath
+from vllm.entrypoints.serve import create_error_response
+from vllm.entrypoints.serve.engine.protocol import (
     ErrorResponse,
     ModelCard,
     ModelList,
     ModelPermission,
 )
-from vllm.entrypoints.openai.models.protocol import BaseModelPath, LoRAModulePath
-from vllm.entrypoints.serve import create_error_response
 from vllm.entrypoints.serve.lora.protocol import (
     LoadLoRAAdapterRequest,
     UnloadLoRAAdapterRequest,

vllm/entrypoints/openai/responses/api_router.py

@@ -8,13 +8,13 @@ from http import HTTPStatus
 from fastapi import APIRouter, Depends, FastAPI, Request
 from fastapi.responses import JSONResponse, StreamingResponse
 
-from vllm.entrypoints.openai.engine.protocol import ErrorResponse
 from vllm.entrypoints.openai.responses.protocol import (
     ResponsesRequest,
     ResponsesResponse,
     StreamingResponsesResponse,
 )
 from vllm.entrypoints.openai.responses.serving import OpenAIServingResponses
+from vllm.entrypoints.serve.engine.protocol import ErrorResponse
 from vllm.entrypoints.serve.utils.api_utils import (
     load_aware_call,
     validate_json_request,

vllm/entrypoints/openai/responses/context.py

@@ -24,11 +24,9 @@ from vllm import envs
 from vllm.entrypoints.chat_utils import (
     ChatTemplateContentFormatOption,
 )
+from vllm.entrypoints.generate.base.protocol import FunctionCall
 from vllm.entrypoints.mcp.tool import Tool
 from vllm.entrypoints.mcp.tool_server import ToolServer
-from vllm.entrypoints.openai.engine.protocol import (
-    FunctionCall,
-)
 from vllm.entrypoints.openai.parser.harmony_utils import render_for_completion
 from vllm.entrypoints.openai.responses.protocol import (
     ResponseInputOutputItem,

vllm/entrypoints/openai/responses/protocol.py

@@ -60,7 +60,8 @@ from vllm.entrypoints.chat_utils import (
     ChatCompletionMessageParam,
     ChatTemplateContentFormatOption,
 )
-from vllm.entrypoints.openai.engine.protocol import OpenAIBaseModel, StopParam
+from vllm.entrypoints.generate.base.protocol import StopParam
+from vllm.entrypoints.serve.engine.protocol import OpenAIBaseModel
 from vllm.exceptions import VLLMValidationError
 from vllm.logger import init_logger
 from vllm.renderers import ChatParams, TokenizeParams, merge_kwargs

vllm/entrypoints/openai/responses/serving.py

@@ -31,16 +31,12 @@ from vllm.entrypoints.chat_utils import (
     ChatCompletionMessageParam,
     ChatTemplateContentFormatOption,
 )
-from vllm.entrypoints.generate.base.serving import (
-    GenerateBaseServing,
-    GenerationError,
-)
-from vllm.entrypoints.mcp.tool_server import ToolServer
-from vllm.entrypoints.openai.engine.protocol import (
+from vllm.entrypoints.generate.base.protocol import (
     DeltaMessage,
-    ErrorResponse,
     RequestResponseMetadata,
 )
+from vllm.entrypoints.generate.base.serving import GenerateBaseServing
+from vllm.entrypoints.mcp.tool_server import ToolServer
 from vllm.entrypoints.openai.models.serving import OpenAIServingModels
 from vllm.entrypoints.openai.parser.harmony_utils import (
     build_harmony_preamble,
@@ -89,9 +85,10 @@ from vllm.entrypoints.openai.responses.utils import (
     extract_function_tool_names,
     extract_tool_types,
 )
+from vllm.entrypoints.serve.engine.protocol import ErrorResponse
 from vllm.entrypoints.serve.utils.api_utils import get_max_tokens
 from vllm.entrypoints.serve.utils.request_logger import RequestLogger
-from vllm.exceptions import VLLMValidationError
+from vllm.exceptions import GenerationError, VLLMValidationError
 from vllm.inputs import EngineInput, tokens_input
 from vllm.logger import init_logger
 from vllm.logprobs import Logprob as SampleLogprob

vllm/entrypoints/openai/responses/streaming_events.py

@@ -61,8 +61,8 @@ from openai.types.responses.response_reasoning_item import (
 from openai.types.responses.tool import Tool
 from openai_harmony import Message as HarmonyMessage
 
+from vllm.entrypoints.generate.base.protocol import DeltaMessage, DeltaToolCall
 from vllm.entrypoints.mcp.tool_server import ToolServer
-from vllm.entrypoints.openai.engine.protocol import DeltaMessage, DeltaToolCall
 from vllm.entrypoints.openai.parser.harmony_utils import (
     extract_function_from_recipient,
     is_function_recipient,

vllm/entrypoints/openai/responses/utils.py

@@ -31,11 +31,11 @@ from openai.types.responses.tool import Tool
 
 from vllm import envs
 from vllm.entrypoints.chat_utils import make_tool_call_id
+from vllm.entrypoints.generate.base.protocol import FunctionCall, FunctionDefinition
 from vllm.entrypoints.openai.chat_completion.protocol import (
     ChatCompletionMessageParam,
     ChatCompletionToolsParam,
 )
-from vllm.entrypoints.openai.engine.protocol import FunctionCall, FunctionDefinition
 from vllm.entrypoints.openai.responses.protocol import ResponseInputOutputItem
 from vllm.logger import init_logger
 from vllm.tool_parsers.utils import (

vllm/entrypoints/pooling/base/protocol.py

@@ -11,7 +11,7 @@ from vllm.entrypoints.chat_utils import (
     ChatCompletionMessageParam,
     ChatTemplateContentFormatOption,
 )
-from vllm.entrypoints.openai.engine.protocol import OpenAIBaseModel
+from vllm.entrypoints.serve.engine.protocol import OpenAIBaseModel
 from vllm.exceptions import VLLMValidationError
 from vllm.renderers import ChatParams, TokenizeParams, merge_kwargs
 from vllm.tasks import check_removed_pooling_task

vllm/entrypoints/pooling/base/serving.py

@@ -16,7 +16,6 @@ from vllm import PoolingRequestOutput, envs
 from vllm.config import VllmConfig
 from vllm.engine.protocol import EngineClient
 from vllm.entrypoints.chat_utils import ChatTemplateConfig
-from vllm.entrypoints.openai.engine.protocol import ErrorResponse
 from vllm.entrypoints.openai.models.serving import OpenAIServingModels
 from vllm.entrypoints.serve.engine.serving import BaseServing
 from vllm.entrypoints.serve.engine.typing import AnyRequest
@@ -30,6 +29,7 @@ from vllm.tracing import (
 )
 from vllm.utils.async_utils import make_async, merge_async_iterators
 
+from ...serve.engine.protocol import ErrorResponse
 from ..typing import AnyPoolingRequest, PoolingServeContext
 from .io_processor import PoolingIOProcessor
 

vllm/entrypoints/pooling/classify/protocol.py

@@ -7,7 +7,7 @@ from typing import Annotated, TypeAlias
 from pydantic import BeforeValidator, Field
 
 from vllm import PoolingParams
-from vllm.entrypoints.openai.engine.protocol import OpenAIBaseModel, UsageInfo
+from vllm.entrypoints.serve.engine.protocol import OpenAIBaseModel, UsageInfo
 from vllm.logger import init_logger
 from vllm.utils import random_uuid
 

vllm/entrypoints/pooling/classify/serving.py

@@ -6,7 +6,7 @@ from typing import TypeAlias
 import numpy as np
 from fastapi.responses import JSONResponse
 
-from vllm.entrypoints.openai.engine.protocol import UsageInfo
+from vllm.entrypoints.serve.engine.protocol import UsageInfo
 from vllm.logger import init_logger
 from vllm.outputs import ClassificationOutput
 

vllm/entrypoints/pooling/embed/api_router.py

@@ -5,7 +5,7 @@ from http import HTTPStatus
 
 from fastapi import APIRouter, Depends, Request
 
-from vllm.entrypoints.openai.engine.protocol import ErrorResponse
+from vllm.entrypoints.serve.engine.protocol import ErrorResponse
 from vllm.entrypoints.serve.utils.api_utils import (
     load_aware_call,
     validate_json_request,

vllm/entrypoints/pooling/embed/protocol.py

@@ -18,7 +18,7 @@ from pydantic import BaseModel, BeforeValidator, Field, model_validator
 
 from vllm import PoolingParams
 from vllm.entrypoints.chat_utils import ChatCompletionMessageParam
-from vllm.entrypoints.openai.engine.protocol import OpenAIBaseModel, UsageInfo
+from vllm.entrypoints.serve.engine.protocol import OpenAIBaseModel, UsageInfo
 from vllm.utils import random_uuid
 
 from ..base.protocol import (

vllm/entrypoints/pooling/pooling/api_router.py

@@ -4,7 +4,7 @@ from http import HTTPStatus
 
 from fastapi import APIRouter, Depends, Request
 
-from vllm.entrypoints.openai.engine.protocol import ErrorResponse
+from vllm.entrypoints.serve.engine.protocol import ErrorResponse
 from vllm.entrypoints.serve.utils.api_utils import (
     load_aware_call,
     validate_json_request,

vllm/entrypoints/pooling/pooling/protocol.py

@@ -7,7 +7,7 @@ from pydantic import BeforeValidator, Field
 
 from vllm import PoolingParams
 from vllm.config import ModelConfig
-from vllm.entrypoints.openai.engine.protocol import OpenAIBaseModel, UsageInfo
+from vllm.entrypoints.serve.engine.protocol import OpenAIBaseModel, UsageInfo
 from vllm.renderers import TokenizeParams
 from vllm.tasks import PoolingTask
 from vllm.utils import random_uuid

vllm/entrypoints/pooling/scoring/api_router.py

@@ -4,7 +4,7 @@ from http import HTTPStatus
 
 from fastapi import APIRouter, Depends, Request
 
-from vllm.entrypoints.openai.engine.protocol import ErrorResponse
+from vllm.entrypoints.serve.engine.protocol import ErrorResponse
 from vllm.entrypoints.serve.utils.api_utils import (
     load_aware_call,
     validate_json_request,

vllm/entrypoints/pooling/scoring/protocol.py

@@ -7,7 +7,7 @@ from pydantic import BaseModel, Field, model_validator
 
 from vllm import PoolingParams
 from vllm.config import ModelConfig
-from vllm.entrypoints.openai.engine.protocol import OpenAIBaseModel, UsageInfo
+from vllm.entrypoints.serve.engine.protocol import OpenAIBaseModel, UsageInfo
 from vllm.renderers import TokenizeParams
 from vllm.tasks import PoolingTask
 from vllm.utils import random_uuid

vllm/entrypoints/pooling/scoring/serving.py

@@ -5,7 +5,7 @@ from fastapi.responses import JSONResponse, Response
 
 from vllm import PoolingParams
 from vllm.engine.protocol import EngineClient
-from vllm.entrypoints.openai.engine.protocol import UsageInfo
+from vllm.entrypoints.serve.engine.protocol import UsageInfo
 from vllm.logger import init_logger
 from vllm.outputs import PoolingRequestOutput, ScoringRequestOutput
 from vllm.tasks import SCORE_TYPE_MAP, SupportedTask

vllm/entrypoints/pooling/utils.py

@@ -14,7 +14,7 @@ import torch
 from fastapi.responses import JSONResponse, StreamingResponse
 
 from vllm.config import ModelConfig
-from vllm.entrypoints.openai.engine.protocol import UsageInfo
+from vllm.entrypoints.serve.engine.protocol import UsageInfo
 from vllm.logger import init_logger
 from vllm.outputs import PoolingRequestOutput
 from vllm.tasks import SupportedTask

vllm/entrypoints/scale_out/derender/api_router.py

@@ -5,7 +5,7 @@ from http import HTTPStatus
 from fastapi import APIRouter, Depends, Request
 from fastapi.responses import JSONResponse
 
-from vllm.entrypoints.openai.engine.protocol import ErrorResponse
+from vllm.entrypoints.serve.engine.protocol import ErrorResponse
 from vllm.entrypoints.serve.utils.api_utils import validate_json_request
 from vllm.logger import init_logger
 

vllm/entrypoints/scale_out/derender/serving.py

@@ -12,14 +12,11 @@ from vllm.entrypoints.openai.completion.protocol import (
     CompletionResponse,
     CompletionStreamResponse,
 )
-from vllm.entrypoints.openai.engine.protocol import (
-    ErrorResponse,
-    UsageInfo,
-)
 from vllm.entrypoints.openai.models.serving import (
     OpenAIModelRegistry,
     OpenAIServingModels,
 )
+from vllm.entrypoints.serve.engine.protocol import ErrorResponse, UsageInfo
 from vllm.entrypoints.serve.engine.serving import BaseServing
 from vllm.entrypoints.serve.utils.request_logger import RequestLogger
 from vllm.inputs import (

vllm/entrypoints/scale_out/render/api_router.py

@@ -8,7 +8,7 @@ from fastapi.responses import JSONResponse
 from vllm.entrypoints.anthropic.protocol import AnthropicMessagesRequest
 from vllm.entrypoints.openai.chat_completion.protocol import ChatCompletionRequest
 from vllm.entrypoints.openai.completion.protocol import CompletionRequest
-from vllm.entrypoints.openai.engine.protocol import ErrorResponse
+from vllm.entrypoints.serve.engine.protocol import ErrorResponse
 from vllm.entrypoints.serve.utils.api_utils import validate_json_request
 from vllm.logger import init_logger