← Models

GPT-5.2-Codex

OpenAI · GPT-5

Current GPT-5.2 coding model for long-horizon software engineering and agentic repository work.

Part of GPT family · Other versions: GPT-5.5 Pro , GPT-5.5 , GPT-Rosalind , GPT-5.4 mini , GPT-5.4 nano , GPT-5.4-Pro , GPT-5.4 , GPT-5.3 , GPT-5.3-Codex , GPT-5.2-Pro , GPT-5.2 , GPT-5-Codex , GPT-5 mini , GPT-5 , GPT-5 nano , GPT-4.1 , GPT-4o mini Transcribe , GPT-4o mini TTS , GPT-4o Transcribe , GPT-4o mini , GPT-4o
Type
language
Context
400K tokens
Max Output
128K tokens
Status
current
Input
$1.75/1M tok
Output
$14/1M tok
API Access
Yes
License
proprietary
coding agentic software-engineering tool-use code-review refactoring
Released December 2025 · Updated April 18, 2026

Overview

Freshness note: Model capabilities, limits, and pricing can change quickly. This profile is a point-in-time snapshot last verified on April 18, 2026.

GPT-5.2-Codex remains a current OpenAI API model. OpenAI’s current model docs still describe it as a coding model optimized for long-horizon, agentic coding tasks. What changed is the surrounding product story: OpenAI’s current Codex clients and help docs no longer support a simple “one model is the default everywhere” narrative.

Capabilities

GPT-5.2-Codex is built for multi-file edits, refactors, test-aware implementation loops, and code review assistance. It tends to perform best when prompts include explicit acceptance criteria, repository conventions, and validation constraints.

Technical Details

OpenAI’s current model card lists GPT-5.2-Codex with a 400K context window, 128K max output, and reasoning effort controls from low through xhigh. It is available through the same modern API surfaces used for tool-calling and structured orchestration, which makes it practical for CI-adjacent agent workflows.

The practical nuance is that this is best read as an API-model snapshot, not as a guaranteed current default inside every Codex client. OpenAI’s own help docs now say Codex model choice depends on the surface, client version, and configuration.

Pricing & Access

Published API pricing (per 1M tokens):

  • Input: $1.75
  • Cached input: $0.175
  • Output: $14.00

This pricing profile is aligned with the current GPT-5.2 flagship tier rather than the older cheaper Codex pricing that appeared in earlier repo snapshots.

Best Use Cases

Use GPT-5.2-Codex for repository-scale implementation, coding agents, and long-running software workflows when you want a current OpenAI coding-specialized model with explicit API documentation and do not need to assume the newest Codex-linked route.

Comparisons

  • GPT-5.3-Codex (OpenAI): Another specialized Codex model still listed by OpenAI for coding-heavy workflows.
  • GPT-5.2 (OpenAI): Better choice when coding is only part of a broader professional workflow.
  • Claude Opus family (Anthropic): Strong alternative for premium coding-heavy workflows outside the OpenAI stack.