TL;DR
- OpenAI released GPT-5.2 in three versions: Instant, Thinking, and Pro.
- The release aims to unlock more economic value and improve capabilities like spreadsheet creation, coding, and image perception.
- GPT-5.2 features a 400,000-token context window and a knowledge cutoff of August 31, 2025.
- The release is a response to competitive pressure from Google's Gemini 3 AI model.
- GPT-5.2 is rolling out to paid ChatGPT subscribers and developers via API.
- Benchmarks show GPT-5.2 Thinking outperforming Gemini 3 Pro and Claude Opus 4.5 on specific software engineering and graduate-level science tasks.
- OpenAI claims GPT-5.2 Thinking beats or ties human professionals on over 70% of tasks in their new GDPval benchmark.
- GPT-5.2 Thinking reportedly generates responses with 38% fewer confabulations than GPT-5.1.
Continue reading
the original article