diff --git a/litellm/llms/bedrock/chat/invoke_handler.py b/litellm/llms/bedrock/chat/invoke_handler.py
index 44e5b40380..9fa791e069 100644
--- a/litellm/llms/bedrock/chat/invoke_handler.py
+++ b/litellm/llms/bedrock/chat/invoke_handler.py
@@ -72,6 +72,9 @@ _response_stream_shape_cache = None
 bedrock_tool_name_mappings: InMemoryCache = InMemoryCache(
     max_size_in_memory=50, default_ttl=600
 )
+from litellm.llms.bedrock.chat.converse_transformation import AmazonConverseConfig
+
+converse_config = AmazonConverseConfig()
 
 
 class AmazonCohereChatConfig:
@@ -1274,7 +1277,7 @@ class AWSEventStreamDecoder:
             text = ""
             tool_use: Optional[ChatCompletionToolCallChunk] = None
             finish_reason = ""
-            usage: Optional[ChatCompletionUsageBlock] = None
+            usage: Optional[Usage] = None
             provider_specific_fields: dict = {}
             reasoning_content: Optional[str] = None
             thinking_blocks: Optional[List[ChatCompletionThinkingBlock]] = None
@@ -1350,12 +1353,7 @@ class AWSEventStreamDecoder:
             elif "stopReason" in chunk_data:
                 finish_reason = map_finish_reason(chunk_data.get("stopReason", "stop"))
             elif "usage" in chunk_data:
-                usage = ChatCompletionUsageBlock(
-                    prompt_tokens=chunk_data.get("inputTokens", 0),
-                    completion_tokens=chunk_data.get("outputTokens", 0),
-                    total_tokens=chunk_data.get("totalTokens", 0),
-                )
-
+                usage = converse_config._transform_usage(chunk_data.get("usage", {}))
             model_response_provider_specific_fields = {}
             if "trace" in chunk_data:
                 trace = chunk_data.get("trace")