AI Coding Tools Roundup: Codex Goes Cloud-Native, GPT-6.1 Sol Reaches Copilot and Four Models Retire

The biggest story for AI coding this week is OpenAI moving Codex fully into the cloud, as reported by TechCrunch. At the same time, GitHub Copilot added GPT-6.1 Sol, shipped a code review API and retired four models. Claude Code and OpenCode had no new stable release since our last roundup.

OpenAI Codex: Reusable Cloud Environments and Security Cloud

At DevDay on September 29, OpenAI announced several Codex changes, summarized by The Decoder:

• Persistent, reusable cloud environments: start a task on your phone, continue on your desktop, and long-running jobs survive a closed laptop.
• A refreshed CLI with voice to start and steer tasks, an /agents view for tracking several tasks, and native worktree support.
• Code review inside the ChatGPT desktop app, including automatic initial reviews while you are away, before feedback goes to GitHub or GitLab.
• Codex Security Cloud, which scans GitHub repositories on a schedule or on every code change, investigates and de-duplicates findings, and prepares fixes for review.
• An Ultrafast tier that speeds token generation up to 8x in Codex at six times standard pricing, limited to Enterprise plans and the new $500/month Pro 500 plan.

On the Codex releases page, the newest builds are still 0.162.0 alpha pre-releases (alpha.14 on October 5) with no written notes, so stay on stable builds for now.

GitHub Copilot: GPT-6.1 Sol, Code Review API and Model Retirements

According to the GitHub changelog, GPT-6.1 Sol is rolling out for agentic coding and terminal workflows to Copilot Pro+, Max, Business and Enterprise. GitHub reports it completed tasks with noticeably fewer tokens and steps than GPT-6 and GPT-5.6. It is available in VS Code, Visual Studio, JetBrains, Xcode, Eclipse, Copilot CLI, github.com and GitHub Mobile, and administrators control access through model policies.

On October 2, GitHub also added API support for Copilot code review: you can now request a review through the REST and GraphQL APIs. The default review effort level changed from Default to Balanced for new and existing repositories and organizations, while users who chose Lite keep it.

The same day, GitHub deprecated four models across all Copilot experiences, with these suggested replacements:

• Gemini 3.5 Flash and Gemini 3.6 Flash: use Gemini 3.8 Flash.
• Kimi K2.7 Code: use Kimi K3.
• Claude Opus 4.7: use Claude Opus 5.5.

Claude Code and OpenCode: No New Stable Releases

Claude Code's latest release is still 2.1.289 from October 3, which fixed permission rule gaps. OpenCode's most recent entry on its changelog is v1.18.34 from September 30, which added session identity headers on model requests and improved macOS binary signing (see also the release notes).

What It Means for Developers

• If your Copilot workflows pin Claude Opus 4.7, Gemini 3.5/3.6 Flash or Kimi K2.7 Code, switch to the suggested replacements now.
• Teams on Copilot Pro+, Max, Business or Enterprise can test GPT-6.1 Sol on agentic and terminal-heavy tasks; admins should check model policies.
• Use the new code review API to trigger Copilot reviews from internal tooling, and verify that Balanced effort suits your repositories.
• Evaluate Codex Security Cloud as a scheduled scanner, but review any generated fix before merging.

Resources & References

• TechCrunch: Codex reusable cloud environments
• The Decoder: OpenAI DevDay Codex and API announcements
• GitHub changelog: GPT-6.1 Sol in GitHub Copilot
• GitHub changelog: Copilot code review API support
• GitHub changelog: Selected models deprecated
• Claude Code releases on GitHub
• OpenAI Codex releases
• OpenCode changelog
• Previous roundup: October 4, 2026