Revise reasoning effort values in OpenAIPromptExecutionSettings by rui-hira · Pull Request #14285 · microsoft/semantic-kernel

Rui Hiraoka (rui-hira) · 2026-08-13T07:49:33Z

Updated the documentation for reasoning effort values to include 'none' and 'xhigh'.

Motivation and Context

The reasoning_effort values currently documented for the GPT-5 series include none, minimal, low, medium, high, xhigh, and max, with support varying by model. The current remarks list only low, medium, high, and minimal, so callers have no indication that none (and others) are valid.
none in particular is not a niche value — it is required for some Chat Completions scenarios on newer models.

Description

This property is object and REST level, none and xhigh are also available.
https://learn.microsoft.com/en-us/rest/api/microsoft-foundry/azureopenai/chat?view=rest-microsoft-foundry-v1#openaireasoningeffort

Contribution Checklist

[x ] The code builds clean without any errors or warnings
[x ] The PR follows the SK Contribution Guidelines and the pre-submission formatting script raises no violations
[ x] All unit tests pass, and I have added new tests where possible
[ x] I didn't break anyone 😄

Updated the documentation for reasoning effort values to include 'none' and 'xhigh'.

Copilot

Pull request overview

Updates the .NET OpenAI connector’s execution settings documentation to expand the documented reasoning_effort values, aiming to reflect newer model/API capabilities.

Changes:

Updated OpenAIPromptExecutionSettings.ReasoningEffort XML remarks to include additional string values ("none", "xhigh").

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

    /// Constrains effort on reasoning for reasoning models.
    /// Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response.
    /// Possible values are:
-    /// <para>- <see cref="string"/> values: <c>"low"</c>, <c>"medium"</c>, <c>"high"</c>, <c>"minimal"</c>;</para>
+    /// <para>- <see cref="string"/> values: <c>"none"</c>, <c>"minimal"</c>, <c>"low"</c>, <c>"medium"</c>, <c>"high"</c>, <c>"xhigh"</c>;</para>
    /// <para>- <see cref="ChatReasoningEffortLevel"/> object;</para>


github-actions

MAF Automated Review — Iteration 1

Result: Findings reported
Scope: full PR (1 commit(s)): ccfd59be36f0
Model: claude-opus-4.8

Overview

This PR revises a single XML-doc line to broaden the documented set of reasoning_effort string values, adding "none" and "xhigh". No executable code, serialization, or API signature changed, so there is no runtime regression and no security impact. However, the changed line now lists these values under <see cref="string"/> values, while the sole string consumer (GetEffortLevel) accepts only low/medium/high/minimal and throws NotSupportedException for anything else — creating a doc-vs-runtime contract mismatch that can steer callers into a request-time exception.

Reviewed the supplied pull-request change set across correctness, security/reliability, architecture, and failure behavior.
1 verified finding remained after source verification (1 medium) across 1 file. Details are attached to the affected lines below.

Affected areas: dotnet/src/Connectors/Connectors.OpenAI/Settings/OpenAIPromptExecutionSettings.cs

github-actions · 2026-08-13T07:56:39Z

    /// Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response.
    /// Possible values are:
-    /// <para>- <see cref="string"/> values: <c>"low"</c>, <c>"medium"</c>, <c>"high"</c>, <c>"minimal"</c>;</para>
+    /// <para>- <see cref="string"/> values: <c>"none"</c>, <c>"minimal"</c>, <c>"low"</c>, <c>"medium"</c>, <c>"high"</c>, <c>"xhigh"</c>;</para>


This line now documents "none" and "xhigh" as accepted string values, but the string-handling path in GetEffortLevel (ClientCore.ChatCompletion.cs:568-578) maps only "low", "medium", "high", and "minimal" and throws NotSupportedException for any other string. A caller who follows this doc and sets ReasoningEffort = "none" (or "xhigh") will get a runtime NotSupportedException when the request is built (same path for OpenAI and Azure OpenAI via AzureClientCore.ChatCompletion.cs:56). Please either restrict the documented string list to the values the switch actually accepts ("minimal", "low", "medium", "high") and note that extended values such as "none"/"xhigh" must be passed via a ChatReasoningEffortLevel object, or extend GetEffortLevel to accept "none"/"xhigh" so the code matches the documented contract.

Rui Hiraoka (rui-hira) · 2026-08-14T05:58:12Z

Need to revise this validation to include current parameters so please do not merge this before the validator is fixed.

semantic-kernel/dotnet/src/Connectors/Connectors.OpenAI/Core/ClientCore.ChatCompletion.cs

Lines 556 to 579 in c028a0c

    
           protected static ChatReasoningEffortLevel? GetEffortLevel(OpenAIPromptExecutionSettings executionSettings) 
        
           { 
        
               var effortLevelObject = executionSettings.ReasoningEffort; 
        
               if (effortLevelObject is null) 
        
               { 
        
                   return null; 
        
               } 
        
               if (effortLevelObject is ChatReasoningEffortLevel effort) 
        
               { 
        
                   return effort; 
        
               } 
        
               if (effortLevelObject is string textEffortLevel) 
        
               { 
        
                   return textEffortLevel.ToUpperInvariant() switch 
        
                   { 
        
                       "LOW" => ChatReasoningEffortLevel.Low, 
        
                       "MEDIUM" => ChatReasoningEffortLevel.Medium, 
        
                       "HIGH" => ChatReasoningEffortLevel.High, 
        
                       "MINIMAL" => new("minimal"), 
        
                       _ => throw new NotSupportedException($"The provided reasoning effort '{textEffortLevel}' is not supported.") 
        
                   }; 
        
               }

Revise reasoning effort values in OpenAIPromptExecutionSettings

ccfd59b

Updated the documentation for reasoning effort values to include 'none' and 'xhigh'.

Rui Hiraoka (rui-hira) requested a review from a team as a code owner August 13, 2026 07:49

Copilot AI lite review requested due to automatic review settings August 13, 2026 07:49

Copilot started reviewing on behalf of Rui Hiraoka (rui-hira) August 13, 2026 07:50 View session

Copilot AI reviewed Aug 13, 2026

View reviewed changes

github-actions Bot reviewed Aug 13, 2026

View reviewed changes

Navigation Menu

Uh oh!

Revise reasoning effort values in OpenAIPromptExecutionSettings - #14285

Revise reasoning effort values in OpenAIPromptExecutionSettings#14285
Rui Hiraoka (rui-hira) wants to merge 1 commit into
microsoft:mainfrom
rui-hira:patch-1

Rui Hiraoka (rui-hira) commented Aug 13, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

github-actions Bot left a comment

Uh oh!

github-actions Bot Aug 13, 2026

Uh oh!

Rui Hiraoka (rui-hira) commented Aug 14, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Uh oh!

Conversation

Rui Hiraoka (rui-hira) commented Aug 13, 2026

Motivation and Context

Description

Contribution Checklist

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

github-actions Bot left a comment

Choose a reason for hiding this comment

MAF Automated Review — Iteration 1

Overview

Uh oh!

github-actions Bot Aug 13, 2026

Choose a reason for hiding this comment

Uh oh!

Rui Hiraoka (rui-hira) commented Aug 14, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants