GPT-5.1-Codex-Max is a specialized large-language model (LLM) optimized for agentic coding tasks—that is, software engineering workflows that involve planning, multi-step reasoning, iteration, refactoring, testing, and working across multiple files and sessions. It builds on the GPT-5.1 / Codex lineage, designed not just for “code autocomplete” but for sustained, project-scale code work.