Reasoning level update #6586
Reasoning level update #6586
Conversation
…n test common/model_presets: always enable gpt-5.1-codex-mini in builtin_model_presets tui: update model selection popup snapshot for gpt-5.1-codex-mini
…ported options tui/app: stop sending Op::OverrideTurnContext on model migration prompt, update config and emit relevant events only
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| ModelPreset { | ||
| id: "gpt-5.1", | ||
| model: "gpt-5.1", | ||
| display_name: "gpt-5.1", | ||
| description: "Broad world knowledge with strong general reasoning.", | ||
| default_reasoning_effort: ReasoningEffort::Medium, | ||
| supported_reasoning_efforts: &[ | ||
| ReasoningEffortPreset { | ||
| effort: ReasoningEffort::None, | ||
| description: "Fastest responses with no reasoning", | ||
| }, | ||
| ReasoningEffortPreset { | ||
| effort: ReasoningEffort::Low, | ||
| description: "Balances speed with some reasoning; useful for straightforward queries and short explanations", | ||
| }, | ||
| ReasoningEffortPreset { | ||
| effort: ReasoningEffort::Medium, | ||
| description: "Provides a solid balance of reasoning depth and latency for general-purpose tasks", | ||
| }, | ||
| ReasoningEffortPreset { | ||
| effort: ReasoningEffort::High, | ||
| description: "Maximizes reasoning depth for complex or ambiguous problems", | ||
| }, |
There was a problem hiding this comment.
The preset in codex-rs/common/src/model_presets.rs now advertises ReasoningEffort::None for gpt‑5.1 (lines 87‑109), replacing the previous Minimal option. The v2 model list contract/tests (codex-rs/app-server/tests/suite/v2/model_list.rs) still assert that gpt‑5.1 returns ReasoningEffort::Minimal, so running the suite now fails because the API response contains none while the fixture expects minimal. Update the model list test (and any downstream schema/docs) to expect the new none value so the API and tests stay in sync.
Useful? React with 👍 / 👎.
# Conflicts: # codex-rs/tui/src/app.rs
| ReasoningEffortConfig::Low => "low", | ||
| ReasoningEffortConfig::Medium => "medium", | ||
| ReasoningEffortConfig::High => "high", | ||
| ReasoningEffortConfig::None => "none", |
There was a problem hiding this comment.
I might be wrong - do we need to add None here?
There was a problem hiding this comment.
Someone will eventually ask for it to try it with the API explicitly. We are not showing none in UI
Automatically update reasoning levels when migrating between models