GitHub Copilot retires six AI models today
Gemini 3.7 Flash, GPT-5.5, GPT-5.4, GPT-5.4 mini, GPT-5 mini and Grok 4.5 stop working across Copilot Chat, inline edits, agent mode and completions from today.
GitHub deprecates six models across every Copilot surface today, 19 October: Gemini 3.7 Flash, GPT-5.5, GPT-5.4, GPT-5.4 mini, GPT-5 mini and Grok 4.5. That covers Copilot Chat, inline edits, ask and agent modes, and code completions, so any workflow, extension or custom agent that names one of these models directly stops working from today rather than degrading gracefully.
GitHub’s suggested replacements are Gemini 3.8 Flash, GPT-5.6 Sol for GPT-5.5 and GPT-5.4, GPT-5.6 Luna for the two mini models, and Grok 4.6. Enterprise and Business customers get the replacements auto-enabled unless an administrator has turned off default model enablement or explicitly disabled them, in which case someone needs to flip that switch in Copilot’s model policies first.
Why it matters: if you’ve hardcoded a model name in a GitHub Action, a Copilot extension config or a custom agent definition, today is when that reference breaks rather than quietly falling back to something else. Worth a quick grep across your repos for the six names above before someone’s agent run fails and you’re debugging it cold.