GitHub will retire six Copilot models across coding workflows on October 19

GitHub says it will deprecate six GitHub Copilot models across Copilot Chat, inline edits, ask and agent modes, and code completions on October 19, 2026. The change affects Gemini 3.7 Flash, GPT-5.5, GPT-5.4, GPT-5.4 mini, GPT-5 mini, and Grok 4.5, according to a September 18 GitHub Changelog notice.

The practical point is simple for engineering teams that have pinned model choices into workflows: check them now, before the cutoff turns a quiet platform update into a broken automation path. GitHub lists replacement options for every retiring model. Gemini 3.7 Flash should move to Gemini 3.8 Flash, GPT-5.5 and GPT-5.4 should move to GPT-5.6 Sol, GPT-5.4 mini and GPT-5 mini should move to GPT-5.6 Luna, and Grok 4.5 should move to Grok 4.6.

Where admins need to look

GitHub says the suggested alternatives are automatically enabled for Copilot Enterprise and Copilot Business customers under default model enablement, unless an administrator has turned off the global default or explicitly disabled the model. That detail matters because many companies treat AI model access as a governance setting, not just a user preference.

For organizations with custom policies, the work is less about clicking a new model in the selector and more about finding where old model names are embedded. Admins should review Copilot settings, internal developer guidance, agent configuration, saved prompts, extension code, and any workflow that assumes a specific model is available. Users will see replacement models in the Copilot Chat model selector once access is enabled in supported Copilot experiences, GitHub says.

CyberOGZ take

This is a reminder that AI coding tools now have a maintenance surface similar to cloud APIs. Model retirement can improve quality, cost, or safety over time, but it also creates operational drift for teams that treat model selection as a stable dependency. The safest rule is to avoid hard coding a model unless the workflow genuinely needs it, then document the owner and fallback path.

No action is required to remove the old models after the deprecation date, according to GitHub. The risk sits earlier, during the migration window, when teams need to confirm that policy choices and automated developer workflows land on supported alternatives.

Sources

• github.blog

Cover photo by Daniil Komov on Pexels, used under the Pexels License.