kerry-berri · Sep 26, 2026 · Sep 26, 2026 · Sep 26, 2026 · Sep 26, 2026 · greptile-apps
diff --git a/ARCHITECTURE.md b/ARCHITECTURE.md
@@ -336,7 +336,7 @@ Each translation is isolated in its own file, making it easy to test and modify
 | `/v1/chat/completions` | Gemini | `llms/gemini/chat/transformation.py` |
 | `/v1/chat/completions` | Vertex AI | `llms/vertex_ai/gemini/transformation.py` |
 | `/v1/chat/completions` | OpenAI | `llms/openai/chat/gpt_transformation.py` |
-| `/v1/messages` (passthrough) | Anthropic | `llms/anthropic/experimental_pass_through/messages/transformation.py` |
+| `/v1/messages` (passthrough) | Anthropic | `llms/anthropic/pass_through/messages/transformation.py` |
 | `/v1/messages` (passthrough) | Bedrock | `llms/bedrock/messages/invoke_transformations/anthropic_claude3_transformation.py` |
 | `/v1/messages` (passthrough) | Vertex AI | `llms/vertex_ai/vertex_ai_partner_models/anthropic/experimental_pass_through/transformation.py` |
 | Passthrough endpoints | All | `proxy/pass_through_endpoints/llm_provider_handlers/` |

diff --git a/litellm/__init__.py b/litellm/__init__.py
@@ -1691,7 +1691,7 @@ def set_global_gitlab_config(config: Dict[str, Any]) -> None:
         SagemakerNovaConfig as SagemakerNovaConfig,
     )
     from .llms.cohere.chat.transformation import CohereChatConfig as CohereChatConfig
-    from .llms.anthropic.experimental_pass_through.messages.transformation import (
+    from .llms.anthropic.pass_through.messages.transformation import (
         AnthropicMessagesConfig as AnthropicMessagesConfig,
     )
     from .llms.bedrock.messages.invoke_transformations.anthropic_claude3_transformation import (

diff --git a/litellm/_lazy_imports_registry.py b/litellm/_lazy_imports_registry.py
@@ -742,7 +742,7 @@
     ),
     "CohereChatConfig": (".llms.cohere.chat.transformation", "CohereChatConfig"),
     "AnthropicMessagesConfig": (
-        ".llms.anthropic.experimental_pass_through.messages.transformation",
+        ".llms.anthropic.pass_through.messages.transformation",
         "AnthropicMessagesConfig",
     ),
     "BedrockClaudePlatformMessagesConfig": (

diff --git a/litellm/caching/caching_handler.py b/litellm/caching/caching_handler.py
@@ -52,7 +52,7 @@
 
 if TYPE_CHECKING:
     from litellm.litellm_core_utils.litellm_logging import Logging as LiteLLMLoggingObj
-    from litellm.llms.anthropic.experimental_pass_through.messages.response_cache import (
+    from litellm.llms.anthropic.pass_through.messages.response_cache import (
         AnthropicMessagesStreamCacheWriter,
     )
     from litellm.types.utils import PromptTokensDetailsWrapper
@@ -127,7 +127,7 @@ def _should_defer_streaming_cache_hit_callbacks(*, cached_result: object) -> boo
     spend and callback records. A plain (non-stream) replay logs here, since nothing
     else will.
     """
-    from litellm.llms.anthropic.experimental_pass_through.messages.response_cache import (
+    from litellm.llms.anthropic.pass_through.messages.response_cache import (
         CachedAnthropicMessagesStreamIterator,
     )
     from litellm.responses.streaming_iterator import BaseResponsesAPIStreamingIterator
@@ -930,7 +930,7 @@ def _convert_cached_result_to_model_response(
         elif (
             call_type == CallTypes.anthropic_messages.value or call_type == CallTypes.aanthropic_messages.value
         ) and isinstance(cached_result, dict):
-            from litellm.llms.anthropic.experimental_pass_through.messages.response_cache import (
+            from litellm.llms.anthropic.pass_through.messages.response_cache import (
                 convert_cached_anthropic_messages_result,
             )
 
@@ -1150,7 +1150,7 @@ def wrap_streaming_result_for_cache(
             return result
         if not isinstance(result, AsyncIterator):
             return result
-        from litellm.llms.anthropic.experimental_pass_through.messages.response_cache import (
+        from litellm.llms.anthropic.pass_through.messages.response_cache import (
             AnthropicMessagesStreamCacheWriter,
         )
 

diff --git a/litellm/integrations/shadow_eval_logger.py b/litellm/integrations/shadow_eval_logger.py
@@ -111,7 +111,7 @@ def _chat_request_from_anthropic_messages(
     because the logged optional_params switch dialect per provider path (the bridge's
     inner completion rewrites them to chat shape mid-flight); the adapter translates
     them alongside the messages, and sampling params copy through untranslated."""
-    from litellm.llms.anthropic.experimental_pass_through.adapters.transformation import (
+    from litellm.llms.anthropic.pass_through.adapters.transformation import (
         LiteLLMAnthropicMessagesAdapter,
     )
 

diff --git a/litellm/integrations/websearch_interception/ARCHITECTURE.md b/litellm/integrations/websearch_interception/ARCHITECTURE.md
@@ -70,7 +70,7 @@ Claude Code (Anthropic's official CLI) sends web search requests using Anthropic
 
 Native tools are converted to LiteLLM standard format **before** sending to the provider:
 
-1. **Conversion Point** (`litellm/llms/anthropic/experimental_pass_through/messages/handler.py`):
+1. **Conversion Point** (`litellm/llms/anthropic/pass_through/messages/handler.py`):
    - In `anthropic_messages()` function (lines 60-127)
    - Runs BEFORE the API request is made
    - Detects native web search tools using `is_web_search_tool()`

diff --git a/litellm/llms/anthropic/chat/guardrail_translation/handler.py b/litellm/llms/anthropic/chat/guardrail_translation/handler.py
@@ -25,7 +25,7 @@
 
 from litellm._logging import verbose_proxy_logger
 from litellm.llms.anthropic.chat.transformation import AnthropicConfig
-from litellm.llms.anthropic.experimental_pass_through.adapters.transformation import (
+from litellm.llms.anthropic.pass_through.adapters.transformation import (
     LiteLLMAnthropicMessagesAdapter,
     is_provider_native_tool_dict,
 )
@@ -365,7 +365,7 @@ def build_stream_error_items(
     def _standalone_block_chunks(self, exc: "ModifyResponseException") -> list[bytes]:
         import uuid
 
-        from litellm.llms.anthropic.experimental_pass_through.messages.fake_stream_iterator import (
+        from litellm.llms.anthropic.pass_through.messages.fake_stream_iterator import (
             FakeAnthropicMessagesStreamIterator,
         )
         from litellm.llms.base_llm.guardrail_translation.utils import (

diff --git a/...imental_pass_through/adapters/__init__.py → ...thropic/pass_through/adapters/__init__.py b/...imental_pass_through/adapters/__init__.py → ...thropic/pass_through/adapters/__init__.py
diff --git a/...rimental_pass_through/adapters/handler.py → ...nthropic/pass_through/adapters/handler.py b/...rimental_pass_through/adapters/handler.py → ...nthropic/pass_through/adapters/handler.py
@@ -11,15 +11,15 @@
 import litellm
 from litellm._logging import verbose_logger
 from litellm.litellm_core_utils.asyncify import run_async_function
-from litellm.llms.anthropic.experimental_pass_through.adapters.transformation import (
+from litellm.llms.anthropic.pass_through.adapters.transformation import (
     AnthropicAdapter,
 )
-from litellm.llms.anthropic.experimental_pass_through.context_management import (
+from litellm.llms.anthropic.pass_through.context_management import (
     AnthropicContextManagementError,
     PolyfillResult,
     apply_context_management,
 )
-from litellm.llms.anthropic.experimental_pass_through.utils import (
+from litellm.llms.anthropic.pass_through.utils import (
     is_reasoning_auto_summary_enabled,
     litellm_logging_obj_from_kwargs,
     local_model_name,
@@ -102,7 +102,7 @@ async def _prepare_context_managed_request(
     user_api_key_auth: "UserAPIKeyAuth | None" = None,
 ) -> PolyfillResult | None:
     """Apply client compaction history, then optional context_management polyfill."""
-    from litellm.llms.anthropic.experimental_pass_through.context_management.editors.compact import (
+    from litellm.llms.anthropic.pass_through.context_management.editors.compact import (
         apply_client_compaction_block_history,
     )
 
@@ -179,7 +179,7 @@ def _polyfill_will_run(
     if edits is None:
         return False
 
-    from litellm.llms.anthropic.experimental_pass_through.context_management.constants import (
+    from litellm.llms.anthropic.pass_through.context_management.constants import (
         COMPACT_EDIT_TYPE,
     )
 
@@ -205,7 +205,7 @@ def _spec_has_non_compact_edits(
     if edits is None:
         return False
 
-    from litellm.llms.anthropic.experimental_pass_through.context_management.constants import (
+    from litellm.llms.anthropic.pass_through.context_management.constants import (
         COMPACT_EDIT_TYPE,
     )
 
@@ -240,7 +240,7 @@ def _normalize_spec_edits(
     if _context_management_explicitly_dropped(additional_drop_params):
         return None
 
-    from litellm.llms.anthropic.experimental_pass_through.context_management.dispatcher import (
+    from litellm.llms.anthropic.pass_through.context_management.dispatcher import (
         _normalize_spec,
     )
 
@@ -437,7 +437,7 @@ def _normalize_reasoning_effort(
         Handles both string ("max") and dict ({"effort": "max", "summary": ...})
         formats. Uses model registry to check supports_xhigh/supports_minimal.
         """
-        from litellm.llms.anthropic.experimental_pass_through.utils import (
+        from litellm.llms.anthropic.pass_through.utils import (
             normalize_reasoning_effort_value,
         )
 

diff --git a/...ss_through/adapters/streaming_iterator.py → ...ss_through/adapters/streaming_iterator.py b/...ss_through/adapters/streaming_iterator.py → ...ss_through/adapters/streaming_iterator.py
@@ -118,7 +118,7 @@ def __init__(self, completion_stream: _ChunkStream):
     @staticmethod
     def _is_combined(chunk: "ModelResponseStream") -> bool:
         """True if ``chunk`` carries response content AND a finish_reason."""
-        from litellm.llms.anthropic.experimental_pass_through.messages.utils import (
+        from litellm.llms.anthropic.pass_through.messages.utils import (
             openai_chat_refusal_text,
         )
 
@@ -1029,7 +1029,7 @@ def _with_refusal_stop_details(
         delta: Final = processed_chunk["delta"]
         if delta.get("stop_reason") == "max_tokens":
             return processed_chunk
-        from litellm.llms.anthropic.experimental_pass_through.messages.utils import (
+        from litellm.llms.anthropic.pass_through.messages.utils import (
             refusal_stop_details,
         )
 
@@ -1083,7 +1083,7 @@ def _delta_has_content(processed_chunk: Mapping[str, object]) -> bool:
     @staticmethod
     def _is_blank_delta(chunk: "ModelResponseStream") -> bool:
         from litellm.llms.anthropic.common_utils import is_empty_unsigned_thinking_block
-        from litellm.llms.anthropic.experimental_pass_through.messages.utils import (
+        from litellm.llms.anthropic.pass_through.messages.utils import (
             openai_chat_refusal_text,
         )
 
@@ -1120,7 +1120,7 @@ def _should_start_new_content_block(self, chunk: "ModelResponseStream") -> bool:
         - Different content types in the response
         - Specific markers in the content
         """
-        from litellm.llms.anthropic.experimental_pass_through.messages.utils import (
+        from litellm.llms.anthropic.pass_through.messages.utils import (
             openai_chat_refusal_text,
         )
 

diff --git a/...l_pass_through/adapters/transformation.py → ...c/pass_through/adapters/transformation.py b/...l_pass_through/adapters/transformation.py → ...c/pass_through/adapters/transformation.py
@@ -8,7 +8,7 @@
 from pydantic import JsonValue, TypeAdapter
 
 import litellm
-from litellm.llms.anthropic.experimental_pass_through.utils import (
+from litellm.llms.anthropic.pass_through.utils import (
     is_reasoning_auto_summary_enabled,
     prompt_cache_key_from_user_id,
 )
@@ -134,14 +134,14 @@ def create_tool_name_mapping(
     normalize_anthropic_tool_use_id,
     strip_encrypted_reasoning_blocks_from_anthropic_messages,
 )
-from litellm.llms.anthropic.experimental_pass_through.context_management import (
+from litellm.llms.anthropic.pass_through.context_management import (
     PolyfillResult,
 )
-from litellm.llms.anthropic.experimental_pass_through.messages.mid_conversation_system import (
+from litellm.llms.anthropic.pass_through.messages.mid_conversation_system import (
     convert_mid_conversation_system_turns,
     is_system_role_message,
 )
-from litellm.llms.anthropic.experimental_pass_through.messages.utils import (
+from litellm.llms.anthropic.pass_through.messages.utils import (
     openai_chat_refusal_text,
     refusal_stop_details,
 )

diff --git a/...experimental_pass_through/architecture.md → ...ms/anthropic/pass_through/architecture.md b/...experimental_pass_through/architecture.md → ...ms/anthropic/pass_through/architecture.md
diff --git a/...ss_through/context_management/__init__.py → ...ss_through/context_management/__init__.py b/...ss_through/context_management/__init__.py → ...ss_through/context_management/__init__.py
diff --git a/...s_through/context_management/constants.py → ...s_through/context_management/constants.py b/...s_through/context_management/constants.py → ...s_through/context_management/constants.py
diff --git a/..._through/context_management/dispatcher.py → ..._through/context_management/dispatcher.py b/..._through/context_management/dispatcher.py → ..._through/context_management/dispatcher.py
diff --git a/...gh/context_management/editors/__init__.py → ...gh/context_management/editors/__init__.py b/...gh/context_management/editors/__init__.py → ...gh/context_management/editors/__init__.py
diff --git a/...ext_management/editors/clear_tool_uses.py → ...ext_management/editors/clear_tool_uses.py b/...ext_management/editors/clear_tool_uses.py → ...ext_management/editors/clear_tool_uses.py
diff --git a/...ugh/context_management/editors/compact.py → ...ugh/context_management/editors/compact.py b/...ugh/context_management/editors/compact.py → ...ugh/context_management/editors/compact.py
@@ -798,7 +798,7 @@ def _count_effective_tokens(
     threshold check matches the downstream ``input_tokens`` metric.
     """
     # Local import to avoid pulling the adapter at module load time.
-    from litellm.llms.anthropic.experimental_pass_through.adapters.transformation import (
+    from litellm.llms.anthropic.pass_through.adapters.transformation import (
         LiteLLMAnthropicMessagesAdapter,
     )
 
@@ -955,7 +955,7 @@ def _build_summary_messages(
     system prompt); the conversation history is translated to OpenAI shape;
     the summarization prompt is appended as a final user turn.
     """
-    from litellm.llms.anthropic.experimental_pass_through.adapters.transformation import (
+    from litellm.llms.anthropic.pass_through.adapters.transformation import (
         LiteLLMAnthropicMessagesAdapter,
     )
 

diff --git a/...pass_through/context_management/errors.py → ...pass_through/context_management/errors.py b/...pass_through/context_management/errors.py → ...pass_through/context_management/errors.py
diff --git a/...hrough/context_management/placeholders.py → ...hrough/context_management/placeholders.py b/...hrough/context_management/placeholders.py → ...hrough/context_management/placeholders.py
diff --git a/...pass_through/context_management/result.py → ...pass_through/context_management/result.py b/...pass_through/context_management/result.py → ...pass_through/context_management/result.py
diff --git a/...gh/messages/agentic_streaming_iterator.py → ...gh/messages/agentic_streaming_iterator.py b/...gh/messages/agentic_streaming_iterator.py → ...gh/messages/agentic_streaming_iterator.py
@@ -336,7 +336,7 @@ async def _settle_task(task: asyncio.Task | None) -> None:
             await task
 
     async def aclose(self) -> None:
-        from litellm.llms.anthropic.experimental_pass_through.messages.streaming_iterator import (
+        from litellm.llms.anthropic.pass_through.messages.streaming_iterator import (
             aclose_if_supported,
         )
 
@@ -379,7 +379,7 @@ async def _process_agentic_hooks(self) -> None:
             if hasattr(result, "__aiter__"):
                 self._follow_up_iterator = result.__aiter__()
             elif isinstance(result, dict):
-                from litellm.llms.anthropic.experimental_pass_through.messages.fake_stream_iterator import (
+                from litellm.llms.anthropic.pass_through.messages.fake_stream_iterator import (
                     FakeAnthropicMessagesStreamIterator,
                 )
                 from litellm.types.llms.anthropic_messages.anthropic_response import (

diff --git a/..._through/messages/fake_stream_iterator.py → ..._through/messages/fake_stream_iterator.py b/..._through/messages/fake_stream_iterator.py → ..._through/messages/fake_stream_iterator.py
diff --git a/...rimental_pass_through/messages/handler.py → ...nthropic/pass_through/messages/handler.py b/...rimental_pass_through/messages/handler.py → ...nthropic/pass_through/messages/handler.py
@@ -215,7 +215,7 @@ async def _try_websearch_short_circuit(
         if response is not None:
             anthropic_response = cast(AnthropicMessagesResponse, response)
             if stream:
-                from litellm.llms.anthropic.experimental_pass_through.messages.fake_stream_iterator import (
+                from litellm.llms.anthropic.pass_through.messages.fake_stream_iterator import (
                     FakeAnthropicMessagesStreamIterator,
                 )
 
@@ -531,7 +531,7 @@ def anthropic_messages_handler(
     # reference the provider cannot resolve. Popped from kwargs so it never reaches the provider.
     skip_mcp_handler: Final = kwargs.pop("_skip_mcp_handler", False)
     if not skip_mcp_handler and tools:
-        from litellm.llms.anthropic.experimental_pass_through.messages.mcp_handler import (
+        from litellm.llms.anthropic.pass_through.messages.mcp_handler import (
             anthropic_messages_with_mcp,
         )
         from litellm.responses.mcp.litellm_proxy_mcp_handler import (

diff --git a/...s_through/messages/interceptors/README.md → ...s_through/messages/interceptors/README.md b/...s_through/messages/interceptors/README.md → ...s_through/messages/interceptors/README.md
diff --git a/...through/messages/interceptors/__init__.py → ...through/messages/interceptors/__init__.py b/...through/messages/interceptors/__init__.py → ...through/messages/interceptors/__init__.py
diff --git a/..._through/messages/interceptors/advisor.py → ..._through/messages/interceptors/advisor.py b/..._through/messages/interceptors/advisor.py → ..._through/messages/interceptors/advisor.py
@@ -67,7 +67,7 @@ async def handle(
         custom_llm_provider: str | None,
         **kwargs,
     ) -> AnthropicMessagesResponse | AsyncIterator:
-        from litellm.llms.anthropic.experimental_pass_through.messages.fake_stream_iterator import (
+        from litellm.llms.anthropic.pass_through.messages.fake_stream_iterator import (
             FakeAnthropicMessagesStreamIterator,
         )
 

diff --git a/...ass_through/messages/interceptors/base.py → ...ass_through/messages/interceptors/base.py b/...ass_through/messages/interceptors/base.py → ...ass_through/messages/interceptors/base.py
diff --git a/...ntal_pass_through/messages/mcp_handler.py → ...opic/pass_through/messages/mcp_handler.py b/...ntal_pass_through/messages/mcp_handler.py → ...opic/pass_through/messages/mcp_handler.py
@@ -180,7 +180,7 @@ async def anthropic_messages_with_mcp(
         )
 
     if stream:
-        from litellm.llms.anthropic.experimental_pass_through.messages.fake_stream_iterator import (
+        from litellm.llms.anthropic.pass_through.messages.fake_stream_iterator import (
             FakeAnthropicMessagesStreamIterator,
         )
 

diff --git a/...rough/messages/mid_conversation_system.py → ...rough/messages/mid_conversation_system.py b/...rough/messages/mid_conversation_system.py → ...rough/messages/mid_conversation_system.py
diff --git a/...l_pass_through/messages/response_cache.py → ...c/pass_through/messages/response_cache.py b/...l_pass_through/messages/response_cache.py → ...c/pass_through/messages/response_cache.py
@@ -6,7 +6,7 @@
 import litellm
 from litellm._logging import verbose_logger
 from litellm.caching.caching_handler import create_cache_write_task
-from litellm.llms.anthropic.experimental_pass_through.messages.streaming_iterator import (
+from litellm.llms.anthropic.pass_through.messages.streaming_iterator import (
     AnthropicMessagesStreamingResponse,
     BaseAnthropicMessagesStreamingIterator,
     _is_message_stop_chunk,

diff --git a/...ss_through/messages/streaming_iterator.py → ...ss_through/messages/streaming_iterator.py b/...ss_through/messages/streaming_iterator.py → ...ss_through/messages/streaming_iterator.py
@@ -16,7 +16,7 @@
 from litellm.litellm_core_utils.litellm_logging import Logging as LiteLLMLoggingObj
 from litellm.litellm_core_utils.logging_worker import GLOBAL_LOGGING_WORKER
 from litellm.llms.anthropic.common_utils import ANTHROPIC_ERROR_STATUS_CODE_MAP
-from litellm.llms.anthropic.experimental_pass_through.messages.utils import INCOMPLETE_STREAM_ERROR_MESSAGE
+from litellm.llms.anthropic.pass_through.messages.utils import INCOMPLETE_STREAM_ERROR_MESSAGE
 from litellm.proxy.pass_through_endpoints.success_handler import (
     PassThroughEndpointLogging,
 )

diff --git a/...l_pass_through/messages/transformation.py → ...c/pass_through/messages/transformation.py b/...l_pass_through/messages/transformation.py → ...c/pass_through/messages/transformation.py
@@ -590,7 +590,7 @@ def get_async_streaming_response_iterator(
         litellm_logging_obj: LiteLLMLoggingObj,
     ) -> AsyncIterator:
         """Helper function to handle Anthropic streaming responses using the existing logging handlers"""
-        from litellm.llms.anthropic.experimental_pass_through.messages.streaming_iterator import (
+        from litellm.llms.anthropic.pass_through.messages.streaming_iterator import (
             BaseAnthropicMessagesStreamingIterator,
         )
 

diff --git a/...perimental_pass_through/messages/utils.py → .../anthropic/pass_through/messages/utils.py b/...perimental_pass_through/messages/utils.py → .../anthropic/pass_through/messages/utils.py
diff --git a/...ss_through/responses_adapters/__init__.py → ...ss_through/responses_adapters/__init__.py b/...ss_through/responses_adapters/__init__.py → ...ss_through/responses_adapters/__init__.py
diff --git a/...ass_through/responses_adapters/handler.py → ...ass_through/responses_adapters/handler.py b/...ass_through/responses_adapters/handler.py → ...ass_through/responses_adapters/handler.py
@@ -109,7 +109,7 @@ def _build_responses_kwargs(
     if isinstance(reasoning, dict):
         effort: Final[object] = reasoning.get("effort")
         if isinstance(effort, str):
-            from litellm.llms.anthropic.experimental_pass_through.utils import (
+            from litellm.llms.anthropic.pass_through.utils import (
                 normalize_reasoning_effort_value,
             )
 

diff --git a/.../responses_adapters/streaming_iterator.py → .../responses_adapters/streaming_iterator.py b/.../responses_adapters/streaming_iterator.py → .../responses_adapters/streaming_iterator.py
@@ -15,7 +15,7 @@
 from litellm.litellm_core_utils.prompt_templates.common_utils import (
     encrypted_reasoning_signature,
 )
-from litellm.llms.anthropic.experimental_pass_through.messages.utils import (
+from litellm.llms.anthropic.pass_through.messages.utils import (
     INCOMPLETE_STREAM_ERROR_MESSAGE,
     refusal_stop_details,
     responses_output_refusal_text,

diff --git a/...ough/responses_adapters/transformation.py → ...ough/responses_adapters/transformation.py b/...ough/responses_adapters/transformation.py → ...ough/responses_adapters/transformation.py
@@ -20,11 +20,11 @@
 from litellm.litellm_core_utils.reasoning_effort_utils import (
     reasoning_effort_from_thinking_budget,
 )
-from litellm.llms.anthropic.experimental_pass_through.messages.utils import (
+from litellm.llms.anthropic.pass_through.messages.utils import (
     refusal_stop_details,
     responses_output_refusal_text,
 )
-from litellm.llms.anthropic.experimental_pass_through.utils import (
+from litellm.llms.anthropic.pass_through.utils import (
     is_reasoning_auto_summary_enabled,
     prompt_cache_key_from_user_id,
 )
@@ -69,7 +69,7 @@ def translate_responses_api_usage_to_anthropic_usage(
         if raw_usage is None:
             return AnthropicUsage(input_tokens=0, output_tokens=0)
 
-        from litellm.llms.anthropic.experimental_pass_through.adapters.transformation import (
+        from litellm.llms.anthropic.pass_through.adapters.transformation import (
             LiteLLMAnthropicMessagesAdapter,
         )
         from litellm.responses.utils import ResponseAPILoggingUtils

diff --git a/...hropic/experimental_pass_through/utils.py → litellm/llms/anthropic/pass_through/utils.py b/...hropic/experimental_pass_through/utils.py → litellm/llms/anthropic/pass_through/utils.py
diff --git a/litellm/llms/anthropic/prompt_cache_prediction.py b/litellm/llms/anthropic/prompt_cache_prediction.py
@@ -16,7 +16,7 @@
 from litellm.llms.anthropic.common_utils import AnthropicModelInfo, is_anthropic_oauth_key
 from litellm.llms.anthropic.count_tokens.handler import AnthropicCountTokensHandler
 from litellm.llms.anthropic.count_tokens.transformation import COUNT_TOKEN_OPTION_NAMES
-from litellm.llms.anthropic.experimental_pass_through.messages.transformation import (
+from litellm.llms.anthropic.pass_through.messages.transformation import (
     DEFAULT_ANTHROPIC_API_VERSION,
     AnthropicMessagesConfig,
 )

diff --git a/litellm/llms/azure_ai/anthropic/messages_transformation.py b/litellm/llms/azure_ai/anthropic/messages_transformation.py
@@ -4,7 +4,7 @@
 
 from typing import Any, Final
 
-from litellm.llms.anthropic.experimental_pass_through.messages.transformation import (
+from litellm.llms.anthropic.pass_through.messages.transformation import (
     AnthropicMessagesConfig,
 )
 from litellm.llms.azure.common_utils import BaseAzureLLM

diff --git a/litellm/llms/bedrock/chat/converse_transformation.py b/litellm/llms/bedrock/chat/converse_transformation.py
@@ -1831,7 +1831,7 @@ def _filter_context_management_for_bedrock_converse(
         anthropic_beta_list: list,
     ) -> None:
         """Keep only compact_20260112 edits for Bedrock; add beta header or drop field."""
-        from litellm.llms.anthropic.experimental_pass_through.context_management.constants import (
+        from litellm.llms.anthropic.pass_through.context_management.constants import (
             COMPACT_EDIT_TYPE,
         )
         from litellm.types.llms.anthropic import ANTHROPIC_BETA_HEADER_VALUES