Symbolon for Copilot

Use your ChatGPT subscription in GitHub Copilot Chat. You sign in with your ChatGPT account, the same login Codex uses, so Plus, Pro, Business and Enterprise plans all work. No API key needed.

Also from the same developer: Lodestar walks you through how code works, explains anything you hover, and reviews your changes on any model in VS Code, including this extension's. 3 free runs included, then $5 one-time.

1. Run Symbolon: Sign In with ChatGPT and sign in in your browser.
2. Pick a model in the Copilot Chat model picker.
3. Set its reasoning effort and speed right in the picker.

Models come from your account's own catalog, in the same format as Codex's models.json, with Codex's published models.json as the fallback. Context windows, reasoning efforts, speed tiers and image support all come from it, so new models show up without an update. Reasoning carries over between turns, and Copilot's context window indicator and prompt caching work as usual.

Requests count toward your Codex usage limits. When you hit one, Symbolon tells you when it resets. Symbolon: Manage shows the signed-in account and plan, and lets you open your usage limits, refresh models, switch accounts, sign out or open the logs. Signing out ends the session with OpenAI.

Development

npm install
npm run compile
npm run lint      # includes a complexity cap of 8
npx -y knip       # unused files, exports and dependencies
node --import ./test/register.mjs test/e2e.mjs path/to/models.json   # end-to-end run against a mocked ChatGPT
npm run package   # builds the .vsix

Unofficial, not affiliated with or endorsed by OpenAI. ChatGPT and Codex are trademarks of OpenAI.