feat: Add "none" reasoning effort support for GPT-5.2 and GPT-5.1#62
Closed
ben-vargas wants to merge 1 commit intonumman-ali:mainfrom
Closed
feat: Add "none" reasoning effort support for GPT-5.2 and GPT-5.1#62ben-vargas wants to merge 1 commit intonumman-ali:mainfrom
ben-vargas wants to merge 1 commit intonumman-ali:mainfrom
Conversation
…T-5.1 Add support for `reasoning_effort: "none"` which disables the reasoning phase entirely, resulting in faster responses when reasoning is not needed. Changes: - Add `gpt-5.2-none` and `gpt-5.1-none` model mappings - Update getReasoningConfig() to allow "none" for GPT-5.1/5.2 general purpose - Codex variants auto-convert "none" to "low" (or "medium" for Mini) - Add 4 new unit tests (197 total) - Update documentation with reasoning effort support matrix References: - OpenAI API docs: "gpt-5.1 defaults to none, supports: none, low, medium, high" - Codex CLI: ReasoningEffort enum includes None variant - Codex CLI: docs/config.md lists "none" as valid config option
numman-ali
pushed a commit
that referenced
this pull request
Dec 17, 2025
…fix, and HTML version update This release combines fixes and features from PRs #62, #63, and #64: ### Added - "none" reasoning effort support for GPT-5.2 and GPT-5.1 general purpose models - gpt-5.2-none and gpt-5.1-none model mappings and presets (now 18 total models) - 4 new unit tests for "none" reasoning behavior (197 total tests) ### Fixed - Orphaned function_call_output 400 API errors - now converts orphans to assistant messages to preserve context while avoiding API errors - OAuth HTML version display updated from 1.0.4 to 4.1.0 ### Technical Details - getReasoningConfig() detects GPT-5.1 general purpose models and allows "none" - Codex variants auto-convert "none" to "low" (or "medium" for Codex Mini) - Orphan handling now works regardless of tools presence in request Contributors: @code-yeongyu (PR #63), @kanemontreuil (PR #64), @ben-vargas (PR #62) 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
4 tasks
numman-ali
added a commit
that referenced
this pull request
Dec 17, 2025
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Adds support for
reasoning_effort: "none"for GPT-5.2 and GPT-5.1 general purpose models, which disables the reasoning phase entirely for faster responses when reasoning is not needed.Motivation
Per OpenAI API documentation and the Codex CLI implementation:
noneand supports:none,low,medium,highnonealong withlow,medium,high,xhighReferences:
ReasoningEffortenum includesNonevariant (codex-rs/protocol/src/openai_models.rs)ReasoningEffort::Nonethrough to API without rejection (codex-rs/core/src/client.rs)model_reasoning_effort = "none"as valid (docs/config.md)Changes
Added
gpt-5.2-noneandgpt-5.1-nonemodel mappings inmodel-map.tssupportsNonedetection ingetReasoningConfig()for GPT-5.1/5.2 general purpose modelsChanged
"low""medium"(unchanged from before)Testing
Manually tested with
openai/gpt-5.2-nonemodel - confirmed working.Reasoning Effort Support Matrix
nonelowmediumhighxhighNotes
"medium"for better coding assistance; users must explicitly configure"none"if desired"none", but we intentionally keep"medium"as our default since this is a coding assistant plugin