Use OpenAI Codex models directly from the GitHub Copilot Chat model picker in Visual Studio Code with your ChatGPT Plus or Pro subscription.
This native VS Code LanguageModelChatProvider handles OpenAI OAuth locally and streams ChatGPT Codex responses into Copilot Chat without a local proxy.
Highlights
ChatGPT OAuth with PKCE, automatic refresh, and VS Code Secret Storage
Multiple named Codex Bridge entries for separate ChatGPT accounts
Live Codex model discovery with six-hour persisted models.dev enrichment
Streaming text, reasoning summaries, images, and agent-mode tool calls
Per-model reasoning effort, summary, Speed Mode, Web Search, and Image Generation controls
Codex-aware context accounting and prompt-cache reuse
Status-bar quota and local inference-token tracking
Browser, manual-callback, and optional Codex CLI session sign-in paths
Connection testing and secret-safe diagnostics
Quick start
Install Codex Bridge for Copilot Chat. You need VS Code 1.125 or newer, GitHub Copilot Chat, and a ChatGPT account with Codex access.
Run Codex Bridge: Add ChatGPT Account, choose a short profile ID such as personal, and complete sign-in. Use manual sign-in if local port 1455 is unavailable.
Open Chat: Manage Language Models, select Add Models → Codex Bridge, name the entry, and enter the same profile ID.
Repeat those steps with another profile ID to keep work and personal ChatGPT accounts available side by side.
The authenticated Codex catalog remains authoritative. Composer controls override workspace defaults; reasoning defaults to High when supported, while hosted Web Search and Image Generation remain off until enabled. Click the Codex status-bar item to inspect five-hour and weekly quota, reset timing, and tokens observed by this extension.
Unofficial project; not affiliated with OpenAI, GitHub, or Microsoft. The ChatGPT Codex backend is not a public compatibility API and can change. Licensed under MIT.
