Frontier AI
Google released gemini-4-argon on the paid Gemini API with a 1M token context, at $2 per million input and $10 per million output tokens.
Why it matters. The introductory price doubles to $4 input and $20 output after November 30, so budget for the higher rate before you commit.
Source: Google blog
The gpt-4o, gpt-4o-mini, gpt-4.1 and gpt-4.1-mini snapshots leave the API on January 14, 2027, and fine-tuned versions follow the same schedule.
Why it matters. OpenAI recommends gpt-6-luna for all four, so start retesting prompts and fine-tunes against it now.
Source: OpenAI API deprecations
Copilot Pro+, Max, Business and Enterprise users can pick Gemini 4 Argon in VS Code, Copilot CLI, the coding agent, JetBrains and other clients.
Why it matters. New models are on by default, so Business and Enterprise admins who want to review it first should set the model policy.
Source: GitHub changelog
Enterprises can call gemini-4-argon through the Gemini Enterprise Agent Platform API in the global region, and use it in Gemini for Business.
Why it matters. Pricing matches the Gemini API, $2 input and $10 output per million tokens until November 30, then $4 and $20.
Source: Google Cloud blog
A terminal conversation over 200K tokens left idle for about 54 minutes is now compacted before its one-hour prompt cache expires.
Why it matters. Returning to a long session no longer reprocesses it all, and idleCompaction set to false in settings.json turns this off.
Source: Claude Code releases