Skip to content

feat(codex): update models and default version#1985

Merged
looplj merged 1 commit into
looplj:unstablefrom
zouxingyuks:feature/codex
Jul 10, 2026
Merged

feat(codex): update models and default version#1985
looplj merged 1 commit into
looplj:unstablefrom
zouxingyuks:feature/codex

Conversation

@zouxingyuks

Copy link
Copy Markdown
Contributor

概述

更新 Codex 渠道的静态模型列表和默认客户端版本,使 AxonHub 能识别新的 GPT-5.6 Codex 模型,并满足上游模型对 Codex 版本的要求。

变更内容

  • codex.DefaultModels() 中新增:
    • gpt-5.6-sol
    • gpt-5.6-terra
    • gpt-5.6-luna
  • 将默认 Codex CLI 版本从 0.142.5 更新为 0.144.1
  • 保持客户端显式传入 Version header 时优先使用客户端值的现有行为。

关联变更

PR #1982 更新的是前端 bundled model catalog;本 PR 更新 Codex transformer 的专用静态模型 registry 和 outbound Version header,两者互补。

验证

  • git diff --check upstream/unstable...HEAD 通过。
  • 已确认当前分支相对最新 upstream/unstable 仅包含 1 个提交和 1 个文件变更。
  • 按仓库 AGENTS.md 约束,未运行 build、lint 或 test。

closes #1984

@greptile-apps

greptile-apps Bot commented Jul 10, 2026

Copy link
Copy Markdown
Contributor

Greptile Summary

This PR updates the Codex transformer defaults. The main changes are:

  • Adds gpt-5.6-sol, gpt-5.6-terra, and gpt-5.6-luna to the static Codex model list.
  • Bumps the fallback Codex CLI Version header from 0.142.5 to 0.144.1.
  • Keeps the constants change scoped to Codex model discovery and default outbound headers.

Confidence Score: 5/5

This looks safe to merge.

  • No blocking issues found in the changed code.

Important Files Changed

Filename Overview
llm/transformer/openai/codex/constants.go Updates Codex static model constants and the fallback Codex version header.

Reviews (1): Last reviewed commit: "feat: Add new Codex models and update de..." | Re-trigger Greptile

@looplj looplj merged commit a7b9d81 into looplj:unstable Jul 10, 2026
5 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

[Feature/功能]: 更新 Codex 模型列表和默认客户端版本

2 participants