OpenAIGPT-5v5.2

GPT-5.2

No ratings yet

Be the first to review this model

The most capable knowledge-work model from OpenAI featuring Auto, Instant, and Thinking modes. Excels at complex analysis, research synthesis, and professional writing.

Released

November 20, 2025

Parameters

Unknown

Context

Pricing

Paid

Benchmark Performance

Benchmark	Category	Score	Performance
MMLU	knowledge	88.0%	88
HumanEval	coding	95.0%	95
MATH	reasoning	98.0%	98
Chatbot Arena ELO	overall	1442	92

Last updated: March 15, 2026

Benchmark scores may vary based on evaluation methodology and conditions.

Reviews(0)

No reviews yet

Be the first to share your experience with GPT-5.2.