OpenAI DevDay 2026: Codex Gets Persistent Cloud Environments, Voice and Security Scanning
At DevDay on 29 September 2026, OpenAI expanded Codex with reusable, persistent cloud environments reachable from any device, a refreshed CLI with voice control and an /agents view, a new code review experience, and Codex Security Cloud for scanning repositories. What was announced and what it signals about where coding agents run.
At its DevDay on 29 September 2026, OpenAI announced a set of upgrades to Codex, its coding agent. The common thread: Codex work moves off your laptop and keeps going without you. (What Is OpenAI Codex?)
What was announced
Reusable cloud environments. Codex can run on your computer, your phone, or in the cloud, and its cloud environments are now persistent and configurable — a shared workspace for a team with approved settings and permissions, so tasks start faster instead of rebuilding from scratch each time.
A refreshed CLI. Start and steer tasks by voice, delegate and track several tasks at once in a new /agents view, edit prompts, resume sessions, work with git worktrees, and a cleaner terminal interface for long sessions. The CLI changelog that week also added an opt-in "instant interrupt" so new input can steer Codex mid-response.
Code review. A new review experience in the ChatGPT desktop app summarises changes and lets you question Codex about potential issues before feedback goes to GitHub pull requests or GitLab merge requests. Automatic reviews can run while you're away. (Review AI-Generated Pull Requests)
Codex Security Cloud. Scans whole GitHub repositories on demand, on a schedule or on new commits, investigates and de-duplicates findings, and prepares fixes in the cloud — even with your laptop closed.
Agents API. OpenAI's Agents API added computer use and can run on Amazon Bedrock Managed Agents.
OpenAI's DevDay recap lists Codex cloud on Plus, Pro, Business, Healthcare, Education and Enterprise; the refreshed CLI and review experience on all plans; and Codex Security Cloud on Pro, Business, Enterprise and Edu. Check the current documentation for your workspace's access and billing: Security Cloud's FAQ explains that scans created after 1 October are billed separately from included plan usage.
The bigger picture
This is the same direction Anthropic (cloud sessions, routines, background agents), Cursor (cloud agents, self-hosted machines, Projects) and Docker (Cloud Sandboxes) have taken: agents need a place to keep working when you leave — and increasingly, that place is persistent rather than a throwaway sandbox per task. (Why AI Coding Agents Need Persistent Workspaces, Docker Cloud Sandboxes)
What it means for you
- Codex users: try cloud environments for long tasks and the review flow for PRs.
- Choosing a tool: the agents are converging on features; model quality on your codebase, pricing limits and where the work runs (and who controls it) are the real differentiators. (Claude Code vs Codex)
Source: OpenAI's DevDay 2026 recap · Codex cloud environments · Codex changelog
EasySpawn gives Claude Code a persistent cloud server of your own — your app, database and files stay put between sessions, reachable from any device. See how it works or join the waitlist.
Related: Best AI Coding Tools · Claude Code Background Sessions · Claude Code Routines · Cursor Projects and Self-Hosted Machines
Keep reading
Replit Can Now Be Driven From Desktop MCP Clients, and Agent Gets Custom API Connectors
In September 2026 Replit made its platform usable from any compatible desktop MCP client — create, find, inspect, update and publish apps from your AI tool of choice (11 September) — and launched custom API connectors in beta for Pro and Enterprise teams (18 September). Plus newer models in Agent. What it enables and what to watch.
MCP's 2026-07-28 Specification: A Stateless Protocol and What Breaks
The Model Context Protocol released its 2026-07-28 specification, the first since November 2025. The core becomes stateless request/response — no initialize handshake, no Mcp-Session-Id — with routing headers, multi round-trip requests, cacheable lists, auth hardening and deprecated Roots, Sampling, Logging and HTTP+SSE. What changes for server builders and users.