Install ProPR, or skip the server
Run npm install -g propr-cli and propr setup on a Linux Docker host (quickstart). ProPR Hosted runs a dedicated instance for you when you would rather not run one.
Codex Cloud gives each task an OpenAI-hosted workspace, works while your laptop sleeps, and opens the pull request when you are ready. ProPR fits when Codex is one agent in a self-hosted, GitHub-centered process.
@codex comments start a cloud chat that pushes a fix to the branch.What you keep: your GitHub repositories and review process, and the ChatGPT plan or OpenAI API key, which ProPR runs through the same Codex.
| Codex Cloud | ProPR |
|---|---|
| Cloud task | Task |
| Cloud environment | Isolated workspace on your server |
@codex review | /review |
@codex comment | Any plain PR comment |
| ChatGPT web, desktop, mobile | Web UI, desktop app, GitHub issue, CLI, MCP |
| OpenAI models only | Codex, plus four other agents per task |
Run npm install -g propr-cli and propr setup on a Linux Docker host (quickstart). ProPR Hosted runs a dedicated instance for you when you would rather not run one.
Open the web UI, sign in with GitHub, and add the repository. Sign in Codex with your ChatGPT plan or OpenAI API key.
Add the AI label to the issue, or run propr plan create from the terminal, then approve the plan.
Review the diff the way you review any pull request. A plain comment on it becomes the next commit on the same branch; /review posts scored findings and /fix applies them. The CLI playbook drives the whole loop from the terminal.
What stays with Codex Cloud: ProPR has no ChatGPT entry point and no diff view before the pull request; the pull request is where you review.
Codex Cloud is OpenAI's own home for background coding tasks. Start a task from the ChatGPT web or desktop app, the mobile app, or the terminal, review the diff, and commit or open a pull request. The GitHub integration adds @codex review, which posts a review that flags P0 and P1 issues, and any other @codex comment, such as a request to fix the CI failures, which starts a cloud chat with the PR as context and can push the fix back to the branch. OpenAI labels that comment-driven flow Codex Cloud (Legacy) and plans to deprecate it. Codex runs OpenAI models only, included in ChatGPT plans from Free to Pro at $200 and above.
ProPR runs Codex on your own server, on your ChatGPT subscription or OpenAI API key, with OpenAI billing you directly. You approve the plan, review the pull request in GitHub, and a plain comment becomes the next commit on the same branch.
One process holds across models, providers, repositories, and maintainers.
| Question | Codex Cloud | ProPR |
|---|---|---|
| OpenAI hosts the coding task | Yes, one workspace per task | No, your server |
| Starts a task from | ChatGPT web, desktop, mobile, CLI, @codex on a PR | Web UI, desktop app, GitHub issues, MCP, CLI |
| Codex implements the change and opens the PR | Yes | Yes |
| GitHub comments drive follow-up commits | @codex comment, Legacy cloud chat | Any plain comment, any agent |
| PR review from a comment | @codex review, P0 and P1 findings | /review scores, /fix applies |
| Switch providers without switching the process | No, OpenAI models only | Yes |
Verified November 2026: Codex Cloud docs, Codex GitHub integration, Codex cloud environments, Codex pricing.