OpenAI drops a new benchmark-topping model, GPT-5.2

AI & Data Science
,
1
TODAY IN AI

670xauto

1. OpenAI drops a new benchmark-topping model, GPT-5.2: OpenAIā€™s newest model is designed to excel at real-world professional tasks and comes in three variants: Instant for quick queries, Thinking for coding and planning, and Pro for your hardest challenges. The standout performer, GPT-5.2 Thinking, matched or beat human experts 70.9% of the time on knowledge work while running 11x faster. The new model is rolling out now, starting with paid plans.