GPT-5.2 Codex (High) — benchmark results

GPT-5.2 Codex evaluated at the high reasoning-effort setting. Provider: OpenAI. Released 2026-01-14. Access: API.

Unified ELO 1877 ± 11, rank #83 of 1776 rated models, from 8 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
APEX v1 Consulting66.9Score (%)100
BinaryAudit88.41Avg Success Rate (%)88
SlopCodeBench21.94Isolated Solved (%)81.2
APEX v1 Investment Banking61.8Score (%)77.8
APEX-Agents42.2Mean Score (ReAct) (self-reported)76.9
APEX v1 Big Law73Score (%)55.6
APEX v1 Medicine (MD)59.7Score (%)50
APEX v165.3Score (%)37.5

Interactive version: theaggregate.ai/model?slug=gpt-5-2-codex-high · How the rankings work · Data refreshed daily, snapshot 2026-07-22.