Merge pull request #37878 from BerriAI/litellm_ruff_no_duplicate_definitions

test: enforce F811 so a duplicate definition cannot silently replace the first

ruff-tests.toml

@@ -26,6 +26,9 @@
 # PT014   the same `parametrize` case listed twice. The copy re-runs an assertion that
 #         already passed and adds no coverage, and it usually marks a case someone meant
 #         to vary and forgot to edit
+# F811    a name bound twice where the first binding was never used. Mostly a repeated
+#         import, but the same rule is what catches a second `def test_x` silently
+#         replacing the first, and a local that shadows an import the module still calls
 #
 # No target-version here on purpose: it resolves from requires-python (>=3.10), so
 # 3.11-only builtins like BaseExceptionGroup are correctly flagged in a tree that
@@ -33,4 +36,4 @@
 
 line-length = 120
 
-lint.select = ["F821", "B011", "B015", "B017", "B018", "PT011", "PT012", "PT014", "PT015", "PLR0133", "PLW0127"]
+lint.select = ["F811", "F821", "B011", "B015", "B017", "B018", "PT011", "PT012", "PT014", "PT015", "PLR0133", "PLW0127"]

tests/audio_tests/test_audio_speech.py

@@ -12,7 +12,6 @@ from litellm._uuid import uuid
 from dotenv import load_dotenv
 
 load_dotenv()
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")
@@ -452,7 +451,7 @@ async def test_azure_ava_tts_with_custom_voice():
     Test that when using a custom Azure voice (en-US-AndrewNeural),
     the SSML request body contains the selected voice.
     """
-    from unittest.mock import AsyncMock, MagicMock, patch
+    from unittest.mock import AsyncMock, patch
     import httpx
 
     # Mock response
@@ -497,7 +496,7 @@ async def test_azure_ava_tts_fable_voice_mapping():
     Test that when using OpenAI voice 'fable',
     it gets mapped to Azure voice 'en-GB-RyanNeural' in the SSML.
     """
-    from unittest.mock import AsyncMock, MagicMock, patch
+    from unittest.mock import AsyncMock, patch
     import httpx
 
     # Mock response
@@ -544,7 +543,7 @@ async def test_aws_polly_tts_with_native_voice():
     Verifies the request is formatted correctly for the Polly API.
     """
     import json
-    from unittest.mock import MagicMock, patch
+    from unittest.mock import patch
     import httpx
 
     # Mock response - Polly returns audio bytes directly
@@ -592,7 +591,7 @@ async def test_aws_polly_tts_with_openai_voice_mapping():
     Verifies that OpenAI voices are correctly mapped to Polly voices.
     """
     import json
-    from unittest.mock import MagicMock, patch
+    from unittest.mock import patch
     import httpx
 
     mock_response_content = b"fake_audio_data"
@@ -634,7 +633,7 @@ async def test_aws_polly_tts_with_ssml():
     Verifies that SSML is detected and TextType is set correctly.
     """
     import json
-    from unittest.mock import MagicMock, patch
+    from unittest.mock import patch
     import httpx
 
     mock_response_content = b"fake_audio_data"

tests/audio_tests/test_whisper.py

@@ -44,7 +44,6 @@ load_dotenv()
 sys.path.insert(
     0, os.path.abspath("../")
 )  # Adds the parent directory to the system path
-import litellm
 from litellm import Router
 
 
@@ -146,7 +145,6 @@ async def test_whisper_log_pre_call():
     from litellm.litellm_core_utils.litellm_logging import Logging
     from datetime import datetime
     from unittest.mock import patch, MagicMock
-    from litellm.integrations.custom_logger import CustomLogger
 
     custom_logger = CustomLogger()
 

tests/enterprise/conftest.py

@@ -35,7 +35,6 @@ def setup_and_teardown():
         0, os.path.abspath("../..")
     )  # Adds the project directory to the system path
 
-    import litellm
     from litellm import Router
 
     importlib.reload(litellm)

tests/enterprise/litellm_enterprise/integrations/test_prometheus.py

@@ -755,7 +755,6 @@ class MockHistogram:
 @pytest.fixture
 def mock_prometheus_logger():
     """Create a PrometheusLogger with mocked metrics to test increment logic"""
-    from unittest.mock import patch
 
     collectors = list(REGISTRY._collector_to_names.keys())
     for collector in collectors:
@@ -1186,7 +1185,7 @@ async def test_langfuse_callback_failure_metric(prometheus_logger):
     This test verifies that when Langfuse logging fails, the 
     litellm_callback_logging_failures_metric is incremented with callback_name="langfuse".
     """
-    from unittest.mock import MagicMock, patch
+    from unittest.mock import MagicMock
 
     from litellm.integrations.langfuse.langfuse_prompt_management import (
         LangfusePromptManagement,
@@ -1242,7 +1241,7 @@ async def test_langfuse_otel_callback_failure_metric(prometheus_logger):
     This test verifies that when Langfuse OTEL logging fails, the 
     litellm_callback_logging_failures_metric is incremented with callback_name="langfuse_otel".
     """
-    from unittest.mock import MagicMock, patch
+    from unittest.mock import MagicMock
 
     from litellm.integrations.langfuse.langfuse_otel import LangfuseOtelLogger
 

tests/enterprise/litellm_enterprise/integrations/test_prometheus_unit_tests.py

@@ -19,7 +19,7 @@ import os
 sys.path.insert(
     0, os.path.abspath("../..")
 )  # Adds the parent directory to the system-path
-from unittest.mock import MagicMock, patch
+from unittest.mock import MagicMock
 
 import pytest
 

tests/guardrails_tests/test_custom_guardrail.py

@@ -26,10 +26,8 @@ from litellm.integrations.custom_guardrail import CustomGuardrail
 
 from typing import Any, Dict, List, Literal, Optional, Union
 
-import litellm
 from litellm._logging import verbose_proxy_logger
 from litellm.caching.caching import DualCache
-from litellm.integrations.custom_guardrail import CustomGuardrail
 from litellm.proxy._types import UserAPIKeyAuth
 from litellm.proxy.guardrails.guardrail_helpers import should_proceed_based_on_metadata
 from litellm.types.guardrails import GuardrailEventHooks

tests/litellm_utils_tests/conftest.py

@@ -42,7 +42,6 @@ def setup_and_teardown():
         0, os.path.abspath("../..")
     )  # Adds the project directory to the system path
 
-    import litellm
 
     importlib.reload(litellm)
 

tests/litellm_utils_tests/test_aws_secret_manager.py

@@ -13,8 +13,6 @@ import litellm.types.utils
 load_dotenv()
 import io
 
-import sys
-import os
 
 # Ensure the project root is in the Python path
 sys.path.insert(0, os.path.abspath(os.path.join(os.path.dirname(__file__), "../../..")))

tests/litellm_utils_tests/test_hashicorp.py

@@ -4,7 +4,6 @@ import pytest
 from dotenv import load_dotenv
 
 load_dotenv()
-import os
 import httpx
 
 sys.path.insert(

tests/litellm_utils_tests/test_health_check.py

@@ -785,19 +785,19 @@ async def test_image_generation_health_check_prompt(monkeypatch):
 
     # Default prompt is used when env var is unset
     monkeypatch.delenv("DEFAULT_HEALTH_CHECK_PROMPT", raising=False)
-    litellm_constants, health_check = reload_modules()
-    health_check_calls = await run_health_check(health_check)
+    reloaded_constants, reloaded_health_check = reload_modules()
+    health_check_calls = await run_health_check(reloaded_health_check)
 
     assert len(health_check_calls) == 1
     assert (
-        health_check_calls[0]["prompt"] == litellm_constants.DEFAULT_HEALTH_CHECK_PROMPT
+        health_check_calls[0]["prompt"] == reloaded_constants.DEFAULT_HEALTH_CHECK_PROMPT
     )
 
     # Environment override should change the prompt without code changes
     override_prompt = "environment override prompt"
     monkeypatch.setenv("DEFAULT_HEALTH_CHECK_PROMPT", override_prompt)
-    litellm_constants, health_check = reload_modules()
-    health_check_calls = await run_health_check(health_check)
+    _, reloaded_health_check = reload_modules()
+    health_check_calls = await run_health_check(reloaded_health_check)
 
     assert len(health_check_calls) == 1
     assert health_check_calls[0]["prompt"] == override_prompt

tests/litellm_utils_tests/test_logging_callback_manager.py

@@ -243,7 +243,7 @@ async def test_slack_alerting_callback_registration(callback_manager):
     from litellm.caching.caching import DualCache
     from litellm.proxy.utils import ProxyLogging
     from litellm.integrations.SlackAlerting.slack_alerting import SlackAlerting
-    from unittest.mock import AsyncMock, patch
+    from unittest.mock import patch
 
     # Mock the async HTTP handler
     with patch(

tests/litellm_utils_tests/test_proxy_budget_reset.py

@@ -8,7 +8,6 @@ import pytest
 from dotenv import load_dotenv
 
 load_dotenv()
-import os
 
 from litellm.proxy._types import LiteLLM_BudgetTableFull
 

tests/litellm_utils_tests/test_secret_manager.py

@@ -9,7 +9,6 @@ from dotenv import load_dotenv
 import json
 
 load_dotenv()
-import os
 import tempfile
 from uuid import uuid4
 

tests/llm_responses_api_testing/base_responses_api.py

@@ -16,7 +16,6 @@ import litellm
 from abc import ABC, abstractmethod
 
 from litellm.integrations.custom_logger import CustomLogger
-import json
 from litellm.types.utils import StandardLoggingPayload
 from litellm.types.llms.openai import (
     ResponseCompletedEvent,

tests/llm_responses_api_testing/conftest.py

@@ -81,7 +81,6 @@ def setup_and_teardown():
         0, os.path.abspath("../..")
     )  # Adds the project directory to the system path
 
-    import litellm
 
     importlib.reload(litellm)
 

tests/llm_responses_api_testing/test_anthropic_responses_api.py

@@ -24,7 +24,6 @@ from litellm.types.llms.openai import (
     ResponseAPIUsage,
     IncompleteDetails,
 )
-import litellm
 from litellm.llms.custom_httpx.http_handler import AsyncHTTPHandler
 from base_responses_api import BaseResponsesAPITest
 from openai.types.responses.function_tool import FunctionTool

tests/llm_responses_api_testing/test_azure_responses_api.py

@@ -52,7 +52,7 @@ async def test_azure_responses_api_status_error():
     Test that 'status' field is not sent in the final request body to Azure API.
     The status field should be filtered out from input messages before making the API call.
     """
-    from unittest.mock import AsyncMock, MagicMock
+    from unittest.mock import MagicMock
     import json
 
     request_data = {
@@ -193,7 +193,6 @@ async def test_azure_responses_api_headers_with_llm_provider_prefix():
     in response._hidden_params["headers"] instead of additional_headers, making them
     accessible via completion.headers in the same way as the completion API.
     """
-    import json
     import httpx
 
     mock_response_data = {

tests/llm_responses_api_testing/test_openai_responses_api.py

@@ -13,7 +13,6 @@ import json
 sys.path.insert(0, os.path.abspath("../.."))
 import litellm
 from litellm.integrations.custom_logger import CustomLogger
-import json
 from litellm.types.utils import StandardLoggingPayload
 from litellm.types.llms.openai import (
     ResponseCompletedEvent,

tests/llm_translation/test_anthropic_completion.py

@@ -14,7 +14,6 @@ from litellm.llms.anthropic.chat import ModelResponseIterator
 
 load_dotenv()
 import io
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")
@@ -360,7 +359,6 @@ def test_process_anthropic_headers_with_no_matching_headers():
 )
 def test_anthropic_tool_use(tool_type, tool_config, message_content):
     """Test Anthropic tool use with computer use and web fetch tools."""
-    from litellm import completion
 
     litellm._turn_on_debug()
 
@@ -951,7 +949,6 @@ def test_anthropic_citations_api():
     """
     Test the citations API
     """
-    from litellm import completion
 
     try:
         resp = completion(
@@ -997,7 +994,6 @@ def test_anthropic_citations_api():
 
 
 def test_anthropic_citations_api_streaming():
-    from litellm import completion
 
     resp = completion(
         model="claude-sonnet-4-5-20250929",
@@ -1044,7 +1040,6 @@ def test_anthropic_citations_api_streaming():
     ],
 )
 def test_anthropic_thinking_output(model):
-    from litellm import completion
 
     litellm._turn_on_debug()
 
@@ -1111,7 +1106,6 @@ def test_anthropic_thinking_output_stream(model):
 
 
 def test_anthropic_custom_headers():
-    from litellm import completion
     from litellm.llms.custom_httpx.http_handler import HTTPHandler
 
     client = HTTPHandler()
@@ -1528,7 +1522,6 @@ def test_anthropic_tool_cache_control():
 
 
 def test_anthropic_streaming():
-    from litellm import completion
 
     request_data = {
         "messages": [

tests/llm_translation/test_azure_ai.py

@@ -19,7 +19,6 @@ from litellm.llms.custom_httpx.http_handler import HTTPHandler
 
 load_dotenv()
 import io
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")

tests/llm_translation/test_azure_openai.py

@@ -255,7 +255,6 @@ def test_get_azure_ad_token_from_username_password(
 
 
 def test_azure_openai_gpt_4o_naming(monkeypatch):
-    from openai import AzureOpenAI
     from pydantic import BaseModel, Field
 
     monkeypatch.setenv("AZURE_API_VERSION", "2024-10-21")
@@ -302,7 +301,6 @@ def test_azure_gpt_4o_with_tool_call_and_response_format(api_version):
     from pydantic import BaseModel
     import litellm
 
-    from openai import AzureOpenAI
 
     client = AzureOpenAI(
         api_key="fake-key",

tests/llm_translation/test_bedrock_agents.py

@@ -8,7 +8,6 @@ import litellm.types
 
 load_dotenv()
 import io
-import os
 import json
 
 sys.path.insert(
@@ -67,7 +66,7 @@ async def test_bedrock_agents_with_streaming():
 
 def test_bedrock_agents_with_custom_params():
     litellm._turn_on_debug()
-    from unittest.mock import MagicMock, patch
+    from unittest.mock import MagicMock
     from litellm.llms.custom_httpx.http_handler import HTTPHandler
 
     client = HTTPHandler()

tests/llm_translation/test_bedrock_completion.py

@@ -13,7 +13,6 @@ import litellm.types
 
 load_dotenv()
 import io
-import os
 import json
 
 sys.path.insert(

tests/llm_translation/test_bedrock_dynamic_auth_params_unit_tests.py

@@ -15,13 +15,7 @@ from litellm.llms.custom_httpx.http_handler import HTTPHandler
 from unittest.mock import Mock
 from litellm.llms.bedrock.base_aws_llm import BaseAWSLLM
 
-import json
-import pytest
-from unittest.mock import patch, Mock
 
-import litellm
-from litellm.llms.custom_httpx.http_handler import HTTPHandler
-from litellm.llms.bedrock.base_aws_llm import BaseAWSLLM
 
 
 def test_bedrock_completion_with_region_name():

tests/llm_translation/test_bedrock_govcloud.py

@@ -475,7 +475,6 @@ class TestBedrockGovCloudSupport:
     @patch("litellm.llms.custom_httpx.http_handler.HTTPHandler.post")
     def test_govcloud_completion_with_cost_tracking(self, mock_post):
         """Test that completion requests with cost tracking use correct pricing for GovCloud models"""
-        from litellm import completion
         from unittest.mock import Mock
         import json
 

tests/llm_translation/test_cohere.py

@@ -6,7 +6,6 @@ from dotenv import load_dotenv
 
 load_dotenv()
 import io
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")
@@ -18,7 +17,6 @@ import pytest
 import litellm
 from litellm import RateLimitError, Timeout, completion, completion_cost, embedding
 from unittest.mock import AsyncMock, patch
-from litellm import RateLimitError, Timeout, completion, completion_cost, embedding
 from litellm.llms.custom_httpx.http_handler import AsyncHTTPHandler
 
 litellm.num_retries = 3

tests/llm_translation/test_infinity.py

@@ -11,11 +11,9 @@ sys.path.insert(
 
 import litellm
 
-import json
 import os
 import sys
-from datetime import datetime
-from unittest.mock import patch, MagicMock, AsyncMock
+from unittest.mock import patch, MagicMock
 
 import pytest
 
@@ -23,7 +21,6 @@ sys.path.insert(
     0, os.path.abspath("../..")
 )  # Adds the parent directory to the system-path
 from test_rerank import assert_response_shape
-import litellm
 
 from base_embedding_unit_tests import BaseLLMEmbeddingTest
 from litellm.llms.custom_httpx.http_handler import HTTPHandler, AsyncHTTPHandler

tests/llm_translation/test_minimax_tts.py

@@ -139,7 +139,6 @@ class TestMinimaxTextToSpeechConfig:
 
         # Mock both litellm.api_key and get_secret_str to return None
         import litellm
-        from unittest.mock import patch
 
         original_api_key = litellm.api_key
         try:
@@ -274,7 +273,6 @@ class TestMinimaxSpeechIntegration:
 
     def test_speech_mock_response(self):
         """Test speech synthesis with mocked response"""
-        from unittest.mock import MagicMock, patch
 
         # Create mock audio data (hex-encoded as MiniMax returns)
         mock_audio_bytes = b"fake audio data for testing"

tests/llm_translation/test_mistral_api.py

@@ -11,7 +11,6 @@ from litellm.llms.anthropic.chat import ModelResponseIterator
 
 load_dotenv()
 import io
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")

tests/llm_translation/test_nvidia_nim.py

@@ -11,13 +11,12 @@ sys.path.insert(
 
 import httpx
 import pytest
-from unittest.mock import patch, MagicMock, AsyncMock
+from unittest.mock import patch, MagicMock
 
 import litellm
 from litellm import Choices, Message, ModelResponse, EmbeddingResponse, Usage
 from litellm import completion
 from base_rerank_unit_tests import BaseLLMRerankTest
-import litellm
 
 
 def test_completion_nvidia_nim():

tests/llm_translation/test_openai_o1.py

@@ -134,7 +134,6 @@ def test_litellm_responses():
     """
     ensures that type of completion_tokens_details is correctly handled / returned
     """
-    from litellm import ModelResponse
     from litellm.types.utils import CompletionTokensDetails
 
     response = ModelResponse(

tests/llm_translation/test_rerank.py

@@ -8,7 +8,6 @@ from dotenv import load_dotenv
 
 load_dotenv()
 import io
-import os
 from typing import Optional, Dict
 
 sys.path.insert(

tests/llm_translation/test_text_completion_unit_tests.py

@@ -6,7 +6,7 @@ from unittest.mock import AsyncMock
 import pytest
 import httpx
 from respx import MockRouter
-from unittest.mock import patch, MagicMock, AsyncMock
+from unittest.mock import patch, MagicMock
 
 sys.path.insert(
     0, os.path.abspath("../..")

tests/llm_translation/test_triton.py

@@ -15,9 +15,7 @@ sys.path.insert(
 import pytest
 import litellm
 
-import pytest
 from litellm.llms.triton.embedding.transformation import TritonEmbeddingConfig
-import litellm
 
 from tests.fake_openai_endpoint import FAKE_OPENAI_API_BASE
 

tests/llm_translation/test_unit_test_bedrock_invoke.py

@@ -9,7 +9,6 @@ import json
 
 load_dotenv()
 import io
-import os
 
 sys.path.insert(0, os.path.abspath("../.."))
 from unittest.mock import AsyncMock, Mock, patch

tests/load_tests/conftest.py

@@ -0,0 +1,5 @@
+from tests.load_tests.memory_leak_utils import (  # noqa: F401  # re-exported so pytest resolves these fixtures by name
+    limit_memory,
+    mock_server,
+    test_router,
+)

tests/load_tests/test_linear_memory_growth.py

@@ -21,12 +21,7 @@ pytest tests/load_tests/test_linear_memory_growth.py -v
 
 import pytest
 
-from tests.load_tests.memory_leak_utils import (
-    limit_memory,  # noqa: F401  # pytest fixture used via dependency injection
-    mock_server,  # noqa: F401  # pytest fixture used via dependency injection
-    run_memory_baseline_test,
-    test_router,  # noqa: F401  # pytest fixture used via dependency injection
-)
+from tests.load_tests.memory_leak_utils import run_memory_baseline_test
 
 # Memory limit for all linear memory growth tests
 MEMORY_LIMIT = "40 MB"

tests/load_tests/test_memory_usage.py

@@ -8,7 +8,6 @@ from dotenv import load_dotenv
 
 load_dotenv()
 import io
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")
@@ -21,13 +20,11 @@ from litellm.router import Router
 from typing import Optional
 from unittest.mock import MagicMock, patch
 
-import asyncio
 import pytest
 import os
 import litellm
 from typing import Callable, Any
 
-import tracemalloc
 import gc
 from typing import Type
 from pydantic import BaseModel

tests/local_testing/cache_unit_tests.py

@@ -9,7 +9,6 @@ from litellm._uuid import uuid
 from dotenv import load_dotenv
 
 load_dotenv()
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")

tests/local_testing/test_acompletion_fallbacks.py

@@ -12,7 +12,6 @@ sys.path.insert(
 import concurrent
 
 from dotenv import load_dotenv
-import asyncio
 import litellm
 
 

tests/local_testing/test_aim_guardrails.py

@@ -463,7 +463,6 @@ async def test_post_call_stream__all_chunks_are_valid(monkeypatch, length: int):
 
 @pytest.mark.asyncio
 async def test_post_call_stream__blocked_chunks(monkeypatch):
-    from litellm.proxy.proxy_server import StreamingCallbackError
 
     init_guardrails_v2(
         all_guardrails=[

tests/local_testing/test_amazing_vertex_completion.py

@@ -6,7 +6,6 @@ from dotenv import load_dotenv
 
 load_dotenv()
 import io
-import os
 
 from test_streaming import streaming_format_tests
 

tests/local_testing/test_anthropic_prompt_caching.py

@@ -7,7 +7,6 @@ from dotenv import load_dotenv
 
 load_dotenv()
 import io
-import os
 
 from test_streaming import streaming_format_tests
 
@@ -210,7 +209,6 @@ def anthropic_messages():
 @pytest.mark.asyncio
 async def test_anthropic_vertex_ai_prompt_caching(anthropic_messages, sync_mode):
     litellm._turn_on_debug()
-    from litellm.llms.custom_httpx.http_handler import HTTPHandler, AsyncHTTPHandler
 
     load_vertex_ai_credentials()
 

tests/local_testing/test_auth_utils.py

@@ -6,7 +6,6 @@ import traceback
 from dotenv import load_dotenv
 
 load_dotenv()
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")

tests/local_testing/test_azure_openai.py

@@ -7,7 +7,6 @@ from dotenv import load_dotenv
 
 load_dotenv()
 import io
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")

tests/local_testing/test_basic_python_version.py

@@ -215,9 +215,7 @@ def test_locked_aiohttp_version_is_not_pool_poisoning():
 
 import os
 import subprocess
-import time
 
-import pytest
 import requests
 
 

tests/local_testing/test_blocked_user_list.py

@@ -14,12 +14,10 @@ from dotenv import load_dotenv
 from fastapi import Request
 
 load_dotenv()
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")
 )  # Adds the parent directory to the system path
-import asyncio
 import logging
 
 import pytest
@@ -57,7 +55,6 @@ verbose_proxy_logger.setLevel(level=logging.DEBUG)
 
 from starlette.datastructures import URL
 
-from litellm.caching.caching import DualCache
 from litellm.proxy._types import (
     BlockUsers,
     DynamoDBArgs,

tests/local_testing/test_braintrust.py

@@ -13,12 +13,10 @@ from dotenv import load_dotenv
 from fastapi import Request
 
 load_dotenv()
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")
 )  # Adds the parent directory to the system path
-import asyncio
 import logging
 from unittest.mock import AsyncMock, MagicMock, patch
 
@@ -29,7 +27,6 @@ from litellm.llms.custom_httpx.http_handler import HTTPHandler
 
 
 def test_braintrust_logging():
-    import litellm
 
     litellm.set_verbose = True
 
@@ -53,7 +50,6 @@ def test_braintrust_logging():
 
 
 def test_braintrust_logging_specific_project_id():
-    import litellm
 
     litellm.set_verbose = True
 

tests/local_testing/test_caching.py

@@ -7,7 +7,6 @@ from litellm._uuid import uuid
 from dotenv import load_dotenv
 
 load_dotenv()
-import os
 import json
 
 sys.path.insert(
@@ -33,7 +32,6 @@ from datetime import timedelta
 messages = [{"role": "user", "content": "who is ishaan Github?  "}]
 # comment
 
-import random
 import string
 
 

tests/local_testing/test_caching_ssl.py

@@ -7,7 +7,6 @@ import traceback
 from dotenv import load_dotenv
 
 load_dotenv()
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")

tests/local_testing/test_completion.py

@@ -7,7 +7,6 @@ from dotenv import load_dotenv
 
 load_dotenv()
 import io
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")
@@ -1380,7 +1379,6 @@ def test_ollama_image():
     """
 
     import base64
-    import io
 
     from PIL import Image
 

tests/local_testing/test_completion_with_retries.py

@@ -3,7 +3,6 @@ import traceback
 from dotenv import load_dotenv
 
 load_dotenv()
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")
@@ -207,7 +206,6 @@ async def test_responses_retry_on_auth_error(sync_mode):
     This validates that the @client decorator properly handles responses/aresponses retries.
     """
     from unittest.mock import patch
-    import openai
 
     num_retries = 2
 

tests/local_testing/test_config.py

@@ -10,7 +10,6 @@ from dotenv import load_dotenv
 
 load_dotenv()
 import io
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")

tests/local_testing/test_cost_calc.py

@@ -6,7 +6,6 @@ from dotenv import load_dotenv
 
 load_dotenv()
 import io
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")

tests/local_testing/test_dual_cache.py

@@ -7,7 +7,6 @@ from litellm._uuid import uuid
 from dotenv import load_dotenv
 
 load_dotenv()
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")

tests/local_testing/test_dynamic_rate_limit_handler.py

@@ -13,7 +13,6 @@ from typing import Optional, Tuple
 from dotenv import load_dotenv
 
 load_dotenv()
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")

tests/local_testing/test_embedding.py

@@ -314,7 +314,6 @@ def test_openai_azure_embedding():
         pytest.fail(f"Error occurred: {e}")
 
 
-from openai.types.embedding import Embedding
 
 
 def _openai_mock_response(*args, **kwargs):
@@ -570,7 +569,6 @@ def test_hf_embedding():
 
 # test_hf_embedding()
 
-from unittest.mock import MagicMock, patch
 
 
 def tgi_mock_post(*args, **kwargs):

tests/local_testing/test_function_call_parsing.py

@@ -8,7 +8,6 @@ from dotenv import load_dotenv
 
 load_dotenv()
 import io
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")

tests/local_testing/test_function_calling.py

@@ -6,7 +6,6 @@ from dotenv import load_dotenv
 
 load_dotenv()
 import io
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")

tests/local_testing/test_function_setup.py

@@ -5,7 +5,7 @@ import traceback
 from dotenv import load_dotenv
 
 load_dotenv()
-import os, io
+import io
 
 sys.path.insert(
     0, os.path.abspath("../..")

tests/local_testing/test_get_optional_params_embeddings.py

@@ -5,7 +5,7 @@ import traceback
 from dotenv import load_dotenv
 
 load_dotenv()
-import os, io
+import io
 
 sys.path.insert(
     0, os.path.abspath("../..")

tests/local_testing/test_helicone_integration.py

@@ -131,7 +131,6 @@ def test_helicone_removes_otel_span_from_metadata():
     to prevent JSON serialization errors.
     """
     from litellm.integrations.helicone import HeliconeLogger
-    from unittest.mock import MagicMock
 
     # Create a mock span object (similar to what OpenTelemetry would create)
     mock_span = MagicMock()

tests/local_testing/test_least_busy_routing.py

@@ -11,7 +11,6 @@ import traceback
 from dotenv import load_dotenv
 
 load_dotenv()
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")

tests/local_testing/test_llm_guard.py

@@ -9,7 +9,6 @@ import traceback
 from dotenv import load_dotenv
 
 load_dotenv()
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")

tests/local_testing/test_lowest_cost_routing.py

@@ -7,7 +7,7 @@ import traceback
 from dotenv import load_dotenv
 
 load_dotenv()
-import os, copy
+import copy
 
 sys.path.insert(
     0, os.path.abspath("../..")

tests/local_testing/test_lowest_latency_routing.py

@@ -13,7 +13,6 @@ from dotenv import load_dotenv
 
 load_dotenv()
 import copy
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")

tests/local_testing/test_ollama.py

@@ -8,7 +8,6 @@ from dotenv import load_dotenv
 
 load_dotenv()
 import io
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")

tests/local_testing/test_openai_moderations_hook.py

@@ -9,7 +9,6 @@ import traceback
 from dotenv import load_dotenv
 
 load_dotenv()
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")
@@ -42,8 +41,6 @@ async def test_openai_moderation_error_raising(monkeypatch):
     user_api_key_dict = UserAPIKeyAuth(api_key=_api_key)
     local_cache = DualCache()
 
-    from litellm.proxy.proxy_server import llm_router
-
     llm_router = litellm.Router(
         model_list=[
             {

tests/local_testing/test_prompt_injection_detection.py

@@ -7,7 +7,6 @@ import traceback
 from dotenv import load_dotenv
 
 load_dotenv()
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")

tests/local_testing/test_pydantic.py

@@ -6,7 +6,6 @@ from dotenv import load_dotenv
 
 load_dotenv()
 import io
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")

tests/local_testing/test_router_budget_limiter.py

@@ -4,7 +4,7 @@ import traceback
 from dotenv import load_dotenv
 
 load_dotenv()
-import os, copy
+import copy
 
 sys.path.insert(
     0, os.path.abspath("../..")

tests/local_testing/test_router_cooldown_handlers.py

@@ -536,7 +536,6 @@ async def test_high_traffic_cooldowns_all_healthy_deployments():
 
     all_deployment_ids = router.get_model_ids()
 
-    import random
     from collections import defaultdict
 
     # Create a defaultdict to track successes and failures for each model ID
@@ -629,7 +628,6 @@ async def test_high_traffic_cooldowns_one_bad_deployment():
 
     all_deployment_ids = router.get_model_ids()
 
-    import random
     from collections import defaultdict
 
     # Create a defaultdict to track successes and failures for each model ID
@@ -727,7 +725,6 @@ async def test_high_traffic_cooldowns_one_rate_limited_deployment():
 
     all_deployment_ids = router.get_model_ids()
 
-    import random
     from collections import defaultdict
 
     # Create a defaultdict to track successes and failures for each model ID

tests/local_testing/test_router_debug_logs.py

@@ -10,7 +10,6 @@ sys.path.insert(
     0, os.path.abspath("../..")
 )  # Adds the parent directory to the system path
 
-import asyncio
 import logging
 
 import litellm

tests/local_testing/test_router_timeout.py

@@ -150,7 +150,6 @@ def test_router_timeout_with_retries_anthropic_model(num_retries, expected_call_
     If request hits custom timeout, ensure it's retried.
     """
     from litellm.llms.custom_httpx.http_handler import HTTPHandler
-    import time
 
     litellm.num_retries = num_retries
     litellm.request_timeout = 0.000001

tests/local_testing/test_sagemaker.py

@@ -7,7 +7,6 @@ from dotenv import load_dotenv
 
 load_dotenv()
 import io
-import os
 import litellm
 from test_streaming import streaming_format_tests
 
@@ -20,7 +19,6 @@ from unittest.mock import AsyncMock, MagicMock, patch
 
 import pytest
 
-import litellm
 from litellm import RateLimitError, Timeout, completion, completion_cost, embedding
 from litellm.llms.custom_httpx.http_handler import AsyncHTTPHandler, HTTPHandler
 from litellm.litellm_core_utils.prompt_templates.factory import anthropic_messages_pt

tests/local_testing/test_secret_detect_hook.py

@@ -15,7 +15,6 @@ from datetime import datetime
 from dotenv import load_dotenv
 
 load_dotenv()
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")
@@ -34,7 +33,6 @@ from litellm_enterprise.enterprise_callbacks.secret_detection import (
 )
 from litellm.proxy.proxy_server import chat_completion
 from litellm.proxy.utils import ProxyLogging, hash_token
-from litellm.router import Router
 
 from tests.fake_openai_endpoint import FAKE_OPENAI_API_BASE
 

tests/local_testing/test_stream_chunk_builder.py

@@ -9,11 +9,11 @@ from typing import List
 from litellm.types.utils import StreamingChoices, ChatCompletionAudioResponse
 
 
-def check_non_streaming_response(completion):
-    assert completion.choices[0].message.audio is not None, "Audio response is missing"
-    print("audio", completion.choices[0].message.audio)
+def check_non_streaming_response(response):
+    assert response.choices[0].message.audio is not None, "Audio response is missing"
+    print("audio", response.choices[0].message.audio)
     assert isinstance(
-        completion.choices[0].message.audio, ChatCompletionAudioResponse
+        response.choices[0].message.audio, ChatCompletionAudioResponse
     ), "Invalid audio response type"
     assert len(completion.choices[0].message.audio.data) > 0, "Audio data is empty"
 
@@ -594,7 +594,6 @@ def test_stream_chunk_builder_multiple_tool_calls():
 
 
 def test_stream_chunk_builder_openai_prompt_caching():
-    from openai import OpenAI
     from pydantic import BaseModel
 
     client = OpenAI(
@@ -639,7 +638,6 @@ def test_stream_chunk_builder_openai_prompt_caching():
 @pytest.mark.flaky(retries=5, delay=2)
 def test_stream_chunk_builder_openai_audio_output_usage():
     from pydantic import BaseModel
-    from openai import OpenAI
     from typing import Optional
 
     client = OpenAI(
@@ -720,7 +718,6 @@ def test_stream_chunk_builder_tool_calls_list():
         Function,
         ModelResponseStream,
         Delta,
-        StreamingChoices,
         ChatCompletionDeltaToolCall,
     )
 

tests/local_testing/test_streaming.py

@@ -951,7 +951,6 @@ def test_vertex_ai_stream(provider):
 
     load_vertex_ai_credentials()
     litellm.set_verbose = True
-    import random
 
     test_models = ["gemini-2.5-flash-lite"]
     for model in test_models:
@@ -2352,7 +2351,6 @@ def test_success_callback_streaming():
 from typing import List, Optional
 
 #### STREAMING + FUNCTION CALLING ###
-from pydantic import BaseModel
 
 
 class Function(BaseModel):
@@ -2569,7 +2567,6 @@ def test_azure_streaming_and_function_calling():
 
 @pytest.mark.asyncio
 async def test_azure_astreaming_and_function_calling():
-    from litellm._uuid import uuid
 
     tools = [
         {

tests/local_testing/test_text_completion.py

@@ -8,7 +8,6 @@ from dotenv import load_dotenv
 
 load_dotenv()
 import io
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")

tests/local_testing/test_tpm_rpm_routing_v2.py

@@ -12,7 +12,6 @@ from typing import Dict
 from dotenv import load_dotenv
 
 load_dotenv()
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")
@@ -399,9 +398,7 @@ async def test_multiple_potential_deployments(sync_mode):
 
 def test_single_deployment_tpm_zero():
     import os
-    from datetime import datetime
 
-    import litellm
 
     model_list = [
         {

tests/local_testing/test_update_spend.py

@@ -14,12 +14,10 @@ from dotenv import load_dotenv
 from fastapi import Request
 
 load_dotenv()
-import os
 
 sys.path.insert(
     0, os.path.abspath("../..")
 )  # Adds the parent directory to the system path
-import asyncio
 import logging
 
 import pytest
@@ -54,7 +52,6 @@ verbose_proxy_logger.setLevel(level=logging.DEBUG)
 
 from starlette.datastructures import URL
 
-from litellm.caching.caching import DualCache
 from litellm.proxy._types import (
     BlockUsers,
     DynamoDBArgs,

tests/logging_callback_tests/test_alerting.py

@@ -19,7 +19,6 @@ from litellm.types.integrations.slack_alerting import AlertType
 # import logging
 # logging.basicConfig(level=logging.DEBUG)
 sys.path.insert(0, os.path.abspath("../.."))
-import asyncio
 import os
 import unittest.mock
 from unittest.mock import AsyncMock, MagicMock, patch
@@ -132,8 +131,6 @@ def test_init():
     print("passed testing slack alerting init")
 
 
-from datetime import datetime, timedelta
-from unittest.mock import AsyncMock, patch
 
 
 @pytest.fixture
@@ -342,7 +339,6 @@ async def test_daily_reports_redis_cache_scheduler():
     # we need this to be 0 so it actualy sends the report
     slack_alerting.alerting_args.daily_report_frequency = 0
 
-    from litellm.router import AlertingConfig
 
     router = litellm.Router(
         model_list=[
@@ -382,7 +378,6 @@ async def test_daily_reports_redis_cache_scheduler():
 @pytest.mark.asyncio
 @pytest.mark.skip(reason="Local test. Test if slack alerts are sent.")
 async def test_send_llm_exception_to_slack():
-    from litellm.router import AlertingConfig
 
     # on async success
     router = litellm.Router(

tests/logging_callback_tests/test_built_in_tools_cost_tracking.py

@@ -9,7 +9,6 @@ from fastapi.routing import APIRoute
 
 load_dotenv()
 import io
-import os
 import time
 import json
 
@@ -102,7 +101,6 @@ async def test_openai_web_search_logging_cost_tracking(
 ):
     """Test web search cost tracking with different search context sizes"""
     test_custom_logger = await _setup_web_search_test()
-    from litellm._uuid import uuid
 
     request_kwargs = {
         "model": "openai/gpt-5-search-api",

tests/logging_callback_tests/test_gcs_pub_sub.py

@@ -15,7 +15,6 @@ from unittest.mock import AsyncMock, patch
 
 import pytest
 
-import litellm
 from litellm import completion
 from litellm._logging import verbose_logger
 from litellm.integrations.gcs_pubsub.pub_sub import *

tests/logging_callback_tests/test_generic_api_callback.py

@@ -16,7 +16,6 @@ from unittest.mock import AsyncMock, patch
 
 import pytest
 
-import litellm
 from litellm import completion
 from litellm._logging import verbose_logger
 from litellm.integrations.gcs_pubsub.pub_sub import *

tests/logging_callback_tests/test_moderations_api_logging.py

@@ -9,7 +9,6 @@ from fastapi.routing import APIRoute
 
 load_dotenv()
 import io
-import os
 import time
 import json
 

tests/logging_callback_tests/test_opentelemetry_unit_tests.py

@@ -9,8 +9,6 @@ import traceback
 from dotenv import load_dotenv
 
 load_dotenv()
-import os
-import asyncio
 
 sys.path.insert(
     0, os.path.abspath("../..")

tests/logging_callback_tests/test_spend_logs.py

@@ -9,7 +9,6 @@ from fastapi.routing import APIRoute
 
 load_dotenv()
 import io
-import os
 import time
 
 # this file is to test litellm/proxy

tests/logging_callback_tests/test_token_counting.py

@@ -9,7 +9,6 @@ from fastapi.routing import APIRoute
 
 load_dotenv()
 import io
-import os
 import time
 import json
 

tests/logging_callback_tests/test_unit_test_litellm_logging.py

@@ -19,8 +19,6 @@ from litellm._service_logger import ServiceLogging
 import asyncio
 
 
-from litellm.litellm_core_utils.litellm_logging import Logging
-import litellm
 
 service_logger = ServiceLogging()
 

tests/logging_callback_tests/test_view_request_resp_logs.py

@@ -10,9 +10,7 @@ import logging
 import tempfile
 from litellm._uuid import uuid
 
-import json
 from datetime import datetime, timedelta, timezone
-from datetime import datetime
 
 import pytest
 

tests/mcp_tests/conftest.py

@@ -33,7 +33,6 @@ def setup_and_teardown():
         0, os.path.abspath("../..")
     )  # Adds the project directory to the system path
 
-    import litellm
     from litellm import Router
 
     importlib.reload(litellm)

tests/mcp_tests/test_mcp_litellm_client.py

@@ -13,7 +13,6 @@ from mcp.client.stdio import stdio_client
 import os
 from litellm import experimental_mcp_client
 import litellm
-import pytest
 import json
 
 

tests/openai_endpoints_tests/test_openai_batches_endpoint.py

@@ -15,7 +15,6 @@ from unittest.mock import patch, MagicMock, AsyncMock
 BASE_URL = "http://localhost:4000"  # Replace with your actual base URL
 API_KEY = "sk-1234"  # Replace with your actual API key
 
-from openai import OpenAI
 
 client = OpenAI(base_url=BASE_URL, api_key=API_KEY)
 

tests/pass_through_unit_tests/test_assemblyai_unit_tests_passthrough.py

@@ -15,20 +15,13 @@ import litellm
 from litellm.litellm_core_utils.litellm_logging import Logging as LiteLLMLoggingObj
 
 
-import json
 import os
 import sys
-from datetime import datetime
-from unittest.mock import AsyncMock, Mock, patch
 
 sys.path.insert(
     0, os.path.abspath("../..")
 )  # Adds the parent directory to the system-path
 
-import httpx
-import pytest
-import litellm
-from litellm.litellm_core_utils.litellm_logging import Logging as LiteLLMLoggingObj
 from litellm.proxy.pass_through_endpoints.llm_provider_handlers.assembly_passthrough_logging_handler import (
     AssemblyAIPassthroughLoggingHandler,
     AssemblyAITranscriptResponse,

tests/pass_through_unit_tests/test_unit_test_passthrough_router.py

@@ -7,7 +7,6 @@ from unittest.mock import AsyncMock, Mock, patch, MagicMock
 sys.path.insert(0, os.path.abspath("../.."))  #
 
 import unittest
-from unittest.mock import patch
 from litellm.proxy.pass_through_endpoints.passthrough_endpoint_router import (
     PassthroughEndpointRouter,
 )

tests/pass_through_unit_tests/test_vertex_ai_live_passthrough.py

@@ -440,9 +440,6 @@ class TestVertexAILivePassthroughIntegration:
 
     def test_vertex_ai_live_route_detection(self):
         """Test that the route detection works correctly"""
-        from litellm.proxy.pass_through_endpoints.success_handler import (
-            PassThroughEndpointLogging,
-        )
 
         handler = PassThroughEndpointLogging()
 
@@ -464,9 +461,6 @@ class TestVertexAILivePassthroughIntegration:
         self, mock_handler_class, mock_logging_obj
     ):
         """Test the success handler integration with Vertex AI Live"""
-        from litellm.proxy.pass_through_endpoints.success_handler import (
-            PassThroughEndpointLogging,
-        )
 
         # Mock the handler
         mock_handler = MagicMock()

tests/proxy_admin_ui_tests/conftest.py

@@ -22,7 +22,6 @@ def setup_and_teardown():
         0, os.path.abspath("../..")
     )  # Adds the project directory to the system path
 
-    import litellm
     from litellm import Router
 
     importlib.reload(litellm)

tests/proxy_admin_ui_tests/test_key_management.py

@@ -12,7 +12,6 @@ from unittest.mock import MagicMock, patch
 
 load_dotenv()
 import io
-import os
 import time
 
 # this file is to test litellm/proxy
@@ -893,9 +892,6 @@ async def test_key_update_with_model_specific_params(prisma_client):
     setattr(litellm.proxy.proxy_server, "master_key", "sk-1234")
     await litellm.proxy.proxy_server.prisma_client.connect()
 
-    from litellm.proxy.management_endpoints.key_management_endpoints import (
-        update_key_fn,
-    )
     from litellm.proxy._types import UpdateKeyRequest
 
     new_key = await generate_key_fn(