← Models

GPT-5.2-Pro

OpenAI · GPT-5

Earlier premium GPT-5.2 tier retained as historical context after GPT-5.4-Pro.

Part of GPT family · Other versions: GPT-5.3-Codex , GPT-5.3 , GPT-5.4-Pro , GPT-5.4 , GPT-5.2-Codex , GPT-5.2 , GPT-5-Codex , GPT-5 nano , GPT-5 mini , GPT-5 , 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
legacy
Input
$15/1M tok
Output
$120/1M tok
API Access
Yes
License
proprietary
frontier reasoning high-accuracy coding analysis tool-use
Released February 2026 · Updated March 6, 2026

Overview

Freshness note: Model capabilities, limits, and pricing can change quickly. This profile is a point-in-time snapshot last verified on February 26, 2026.

GPT-5.2-Pro was OpenAI’s premium GPT-5.2 tier for teams that prioritized top-end quality over token cost. It now sits behind GPT-5.4-Pro in the current OpenAI lineup, but it remains useful as historical context for premium GPT-5 pricing and routing.

Capabilities

GPT-5.2-Pro is suited for complex analytical tasks, difficult engineering problem-solving, and high-stakes decision-support workflows where deeper reasoning and output quality are prioritized above latency or throughput economics.

Technical Details

OpenAI’s model listing shows GPT-5.2-Pro in the same 400K context and 128K output class as the rest of the GPT-5.2 line. It is designed as the top quality tier in that generation rather than a cost-optimized variant.

Pricing & Access

Published API pricing (per 1M tokens):

  • Input: $15.00
  • Output: $120.00

This price band makes GPT-5.2-Pro best reserved for selective routing on high-value tasks.

Best Use Cases

Use this page as a historical reference. For current high-stakes routing, prefer GPT-5.4-Pro unless you have a compatibility reason to stay on older GPT-5.2-era behavior.

Comparisons

  • GPT-5.4-Pro (OpenAI): Current premium successor in the GPT family.
  • GPT-5.4 (OpenAI): Better current default unless the highest-capability tier is required.
  • Claude Opus 4.6 (Anthropic): Contemporary premium alternative for high-quality reasoning work.