Python: add(azure-ai): support reasoning config for AzureAIClient (#3403)

* add(azure-ai): support reasoning config for AzureAIClient

* Update sample

* Merge main

* improvements

* improve sample

python/packages/azure-ai/agent_framework_azure_ai/_client.py

@@ -22,12 +22,7 @@ from agent_framework.observability import use_instrumentation
 from agent_framework.openai import OpenAIResponsesOptions
 from agent_framework.openai._responses_client import OpenAIBaseResponsesClient
 from azure.ai.projects.aio import AIProjectClient
-from azure.ai.projects.models import (
-    MCPTool,
-    PromptAgentDefinition,
-    PromptAgentDefinitionText,
-    RaiConfig,
-)
+from azure.ai.projects.models import MCPTool, PromptAgentDefinition, PromptAgentDefinitionText, RaiConfig, Reasoning
 from azure.core.credentials_async import AsyncTokenCredential
 from azure.core.exceptions import ResourceNotFoundError
 from pydantic import ValidationError
@@ -51,12 +46,15 @@ else:
 logger = get_logger("agent_framework.azure")
 
 
-class AzureAIProjectAgentOptions(OpenAIResponsesOptions):
+class AzureAIProjectAgentOptions(OpenAIResponsesOptions, total=False):
     """Azure AI Project Agent options."""
 
     rai_config: RaiConfig
     """Configuration for Responsible AI (RAI) content filtering and safety features."""
 
+    reasoning: Reasoning  # type: ignore[misc]
+    """Configuration for enabling reasoning capabilities (requires azure.ai.projects.models.Reasoning)."""
+
 
 TAzureAIClientOptions = TypeVar(
     "TAzureAIClientOptions",
@@ -343,6 +341,10 @@ class AzureAIClient(OpenAIBaseResponsesClient[TAzureAIClientOptions], Generic[TA
                 args["temperature"] = run_options["temperature"]
             if "top_p" in run_options:
                 args["top_p"] = run_options["top_p"]
+            if "reasoning" in run_options:
+                args["reasoning"] = run_options["reasoning"]
+            if "rai_config" in run_options:
+                args["rai_config"] = run_options["rai_config"]
 
             # response_format is accessed from chat_options or additional_properties
             # since the base class excludes it from run_options
@@ -408,6 +410,7 @@ class AzureAIClient(OpenAIBaseResponsesClient[TAzureAIClientOptions], Generic[TA
             "top_p",
             "text",
             "text_format",
+            "reasoning",
         ]
 
         for property in exclude:

python/packages/azure-ai/agent_framework_azure_ai/_project_provider.py

@@ -195,6 +195,7 @@ class AzureAIProjectAgentProvider(Generic[TOptions_co]):
         opts = dict(default_options) if default_options else {}
         response_format = opts.get("response_format")
         rai_config = opts.get("rai_config")
+        reasoning = opts.get("reasoning")
 
         args: dict[str, Any] = {"model": resolved_model}
 
@@ -206,6 +207,8 @@ class AzureAIProjectAgentProvider(Generic[TOptions_co]):
             )
         if rai_config:
             args["rai_config"] = rai_config
+        if reasoning:
+            args["reasoning"] = reasoning
 
         # Normalize tools and separate MCP tools from other tools
         normalized_tools = normalize_tools(tools)