GPT-5.2 Pro

Released 2025-12 · thinking · 400K tokens · 2 benchmarks

Editorial notes

Variante Pro lanzada 11 diciembre 2025 con xhigh effort (5to nivel de reasoning). 42% mejora vs GPT-5.1 en code static analysis. Pricing $1.75/$14 per MTok (max output 128K). Knowledge cutoff 31 agosto 2025.

Spec sheet

Company: OpenAI
Country: US
Type: thinking
Release: 2025-12
Context: 400K tokens
Slug: gpt-5-2-pro

Benchmarks (2)

Reasoning 1

GPQA-Diamond

Graduate-level Physics, Chemistry, Biology — PhD-level questions.

93.2

Agentic 1

GDPval

Real economic-value tasks (real productivity).

70.9

Cite this model

BibTeX · APA

BibTeX

@misc{frontier-gpt-5-2-pro,
  title  = {GPT-5.2 Pro},
  author = {{OpenAI}},
  year   = {2025},
  note   = {Frontier Benchmarks AI atlas. Accessed 2026-05-08},
  url    = {https://frontierbenchmarks.com/models/gpt-5-2-pro}
}

APA

OpenAI (2025). GPT-5.2 Pro [Large language model]. Frontier Benchmarks AI. Retrieved 2026-05-08, from https://frontierbenchmarks.com/models/gpt-5-2-pro

Citation reflects the atlas page, not the original model paper. For the paper, see the "Resources" section above.

⚔️ Battle vs another model ← All models More from OpenAI Methodology