AWS Security ChangesHomeSearch

AWS bedrock documentation change

Service: bedrock · 2026-04-22 · Documentation low

File: bedrock/latest/userguide/claude-messages-adaptive-thinking.md

Summary

Added documentation for using adaptive thinking with the Converse API, including code examples for the thinking and effort parameters.

Security assessment

The change is purely a feature documentation update for the Converse API, providing code examples for adaptive thinking functionality. There is no mention of security features, vulnerabilities, or security-related changes in the diff.

Diff

diff --git a/bedrock/latest/userguide/claude-messages-adaptive-thinking.md b/bedrock/latest/userguide/claude-messages-adaptive-thinking.md
index de5b49a3e..c94d7cc51 100644
--- a//bedrock/latest/userguide/claude-messages-adaptive-thinking.md
+++ b//bedrock/latest/userguide/claude-messages-adaptive-thinking.md
@@ -7 +7 @@
-How adaptive thinking worksAdaptive thinking with the effort parameterPrompt cachingTuning thinking behavior
+How adaptive thinking worksAdaptive thinking with the effort parameterUsing adaptive thinking with the Converse APIPrompt cachingTuning thinking behavior
@@ -139,0 +140,41 @@ Effort level | Thinking behavior
+## Using adaptive thinking with the Converse API
+
+When using the [Converse API](./conversation-inference.html), pass the `thinking` and `effort` parameters inside `additionalModelRequestFields`. The following example shows adaptive thinking with the default effort level:
+    
+    
+    import boto3, json
+    
+    bedrock_runtime = boto3.client(service_name='bedrock-runtime', region_name='us-east-2')
+    
+    response = bedrock_runtime.converse(
+        modelId="us.anthropic.claude-opus-4-6-v1",
+        messages=[{
+            "role": "user",
+            "content": [{"text": "Explain why the sum of two even numbers is always even."}]
+        }],
+        additionalModelRequestFields={
+            "thinking": {
+                "type": "adaptive"
+            }
+        }
+    )
+    
+    print(json.dumps(response["output"], indent=2, default=str))
+
+To specify an effort level, add the `effort` field inside the `thinking` object:
+    
+    
+    response = bedrock_runtime.converse(
+        modelId="us.anthropic.claude-opus-4-6-v1",
+        messages=[{
+            "role": "user",
+            "content": [{"text": "What is 2 + 2?"}]
+        }],
+        additionalModelRequestFields={
+            "thinking": {
+                "type": "adaptive",
+                "effort": "low"
+            }
+        }
+    )
+