GPT-5.2-Codex
Agentic coding model for complex software engineering and defensive cybersecurity.
About
Released December 18, 2025, GPT-5.2-Codex is OpenAI's agentic coding model tuned for complex software engineering work and defensive cybersecurity tasks. It sits in the Codex lineage of cloud-based software engineering agents (Codex launched May 2025), optimized for long-running autonomous coding sessions rather than chat. It was followed by GPT-5.3-Codex on February 5, 2026.
Limitations: superseded within two months by GPT-5.3-Codex, so it is a single checkpoint rather than a maintained line. As a specialized agentic coder it is not the general-purpose choice; frontier chat/reasoning workloads belong on the GPT-5.x Thinking/Pro tiers. Defensive-cybersecurity framing means offensive security capabilities are restricted.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
No public benchmark numbers located; positioned for agentic coding evals
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 7/10
Scored 7/10 (metrics tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 56; Trending score 72/100 from researcher assessment.