diff --git a/litellm/llms/bedrock/chat/converse_transformation.py b/litellm/llms/bedrock/chat/converse_transformation.py
index b995a948dd8..a4d37810f82 100644
--- a/litellm/llms/bedrock/chat/converse_transformation.py
+++ b/litellm/llms/bedrock/chat/converse_transformation.py
@@ -426,8 +426,8 @@ class AmazonConverseConfig(BaseConfig):
         if "gpt-oss" in model:
             optional_params["reasoning_effort"] = reasoning_effort
         elif "openai.gpt-5" in model:
-            optional_params.pop("thinking", None)
-            optional_params["reasoning"] = {"effort": reasoning_effort}
+            reasoning: Final[BedrockConverseGptReasoningEffortBlock] = {"effort": reasoning_effort}
+            optional_params["reasoning"] = reasoning
         elif self._is_nova_2_model(model):
             reasoning_config: Final = self._transform_reasoning_effort_to_reasoning_config(reasoning_effort)
             optional_params.update(reasoning_config)
diff --git a/litellm/types/llms/bedrock.py b/litellm/types/llms/bedrock.py
index 6ae2e31fe60..ed23db597ae 100644
--- a/litellm/types/llms/bedrock.py
+++ b/litellm/types/llms/bedrock.py
@@ -3,7 +3,7 @@ from collections.abc import Sequence
 from enum import Enum
 from typing import TYPE_CHECKING, Any, Final, Literal
 
-from typing_extensions import Required, TypedDict, override
+from typing_extensions import ReadOnly, Required, TypedDict, override
 
 from .openai import ChatCompletionToolCallChunk
 
@@ -97,6 +97,10 @@ class BedrockConverseReasoningContentBlockDelta(TypedDict, total=False):
     text: str
 
 
+class BedrockConverseGptReasoningEffortBlock(TypedDict):
+    effort: ReadOnly[str]
+
+
 class GuardrailConverseTextBlock(TypedDict, total=False):
     text: str
 
diff --git a/tests/test_litellm/llms/bedrock/test_cross_region_inference_profile_mapping.py b/tests/test_litellm/llms/bedrock/test_cross_region_inference_profile_mapping.py
index dbd31c7e81b..1388073381e 100644
--- a/tests/test_litellm/llms/bedrock/test_cross_region_inference_profile_mapping.py
+++ b/tests/test_litellm/llms/bedrock/test_cross_region_inference_profile_mapping.py
@@ -293,15 +293,16 @@ def test_bedrock_gpt_5_6_advertises_only_converse_supported_features(
 
 
 @pytest.mark.parametrize("profile", GPT_5_6_PROFILES, ids=lambda p: p.model_id)
-def test_bedrock_gpt_5_6_offers_tools_but_not_reasoning(profile, local_model_cost_map):
-    """Converse rejects the Anthropic-shaped thinking block LiteLLM emits for
-    reasoning_effort, so neither reasoning param may be offered yet, while the tool
-    params these models do accept must be."""
+def test_bedrock_gpt_5_6_offers_tools_and_reasoning_effort_but_not_thinking(profile, local_model_cost_map):
+    """GPT-5.x on Converse maps reasoning_effort to reasoning.effort, so reasoning_effort
+    is offered while the Anthropic-only thinking/output_config are not, alongside the tool
+    params these models accept."""
     supported = AmazonConverseConfig().get_supported_openai_params(
         model=f"bedrock/{profile.model_id}"
     )
 
     assert "tools" in supported
     assert "tool_choice" in supported
-    assert "reasoning_effort" not in supported
+    assert "reasoning_effort" in supported
     assert "thinking" not in supported
+    assert "output_config" not in supported