OpenAI Counters Gemini 3 Launch with GPT-5.1-Codex-Max Coding Model
SMRTR summary
OpenAI launched GPT-5.1-Codex-Max just 24 hours after Google's Gemini 3 release, introducing "compaction" technology that enables 24-hour autonomous coding sessions by intelligently managing memory instead of expanding context windows. The model achieves 77.9% on SWE-bench Verified while using 30% fewer tokens, directly countering Gemini 3's performance claims.
SMRTR provides this summary for quick context. The original article belongs to WinBuzzer.
Read the original article