OpenAI expands Codex to mobile, remote SSH and enterprise tools amid Anthropic restrictions

OpenAI has broadened the reach of its Codex code‑generation model by embedding it into the ChatGPT mobile app, enabling users to initiate tasks, review outputs, approve commands and steer execution from their phones while the model continues running on a laptop, Mac mini or dedicated devbox. The company also announced that Remote SSH is now generally available for managed remote environments, and added hooks and programmatic access tokens to support business and enterprise automation around the Codex loop. These moves position Codex as a more flexible, cloud‑agnostic tool that can be integrated into existing development workflows.

In parallel, GitHub’s technical preview of the Copilot App introduces a desktop environment that supports parallel workstreams, repository and pull‑request lifecycle management, and model flexibility. Visual Studio Code has shipped a new Agents window that supports multi‑agent, multi‑project workflows, browser and mobile support via vscode.dev/agents, bring‑your‑own‑key (BYOK) improvements, and terminal output compression. Open‑source projects such as Nous/Hermes Agent and Kimi’s Web Bridge browser extension have added Codex runtime integration, routing OpenAI turns through a Codex CLI or app‑server, and exposing web interaction to multiple code CLIs.

The announcement comes amid Anthropic’s recent restrictions on Claude Code usage for third‑party wrappers and high‑volume programmatic workflows. Those limits have triggered dramatic rate‑limit reductions and subscription cancellations, underscoring the instability of subscription‑backed harnesses and the growing need for provider or model abstraction and BYOK paths. OpenAI’s expansion therefore not only broadens its own product suite but also responds to a market that increasingly demands flexible, self‑hosted or hybrid code‑generation solutions.