OpenAI Codex CLI Adds Google Cloud Vertex AI Support

What happened

The latest update to the OpenAI Codex CLI introduces a built-in Google Cloud Vertex AI provider, which can be configured using the parameter model_provider = "gcp". This addition expands the tool's capabilities by integrating with Google Cloud services.

Additionally, the release includes GPT-6.1 Terra, a new model added to the Codex CLI's model catalog. The update also enhances the user interface with a repository filter in the session picker and directory grouping in the agent command center.

Why it matters

These changes improve usability and functionality for developers interacting with AI models. The integration with Google Cloud Vertex AI allows users to leverage cloud-based AI services, potentially increasing efficiency and scalability in their workflows.

Key details

The app-server API now supports experimental attached thread forks, which may enhance concurrent processing capabilities. Memory v2 is the default for fresh profiles, and support for importing external-agent memories has been added.

TUI preferences are now editable and can be applied immediately, even with invalid execution settings. Remote MCP tool requests are canceled when their local response futures are dropped, improving system reliability.