OpenAI (OPENAI) has launched GPT-5.2, its latest flagship large language model, just one month following the release of GPT-5.1, as the San Francisco-based startup appears determined to stay ahead in the AI race.
GPT‑5.2 Instant, Thinking and Pro models start rolling out today for ChatGPT users, starting with paid plans. They are all available now in the API for developers. Instant is faster at writing and information seeking; Thinking is designed for structured work like coding and planning; while Pro delivers the most accurate answers for complex questions, according to OpenAI.
“GPT-5.2 is here,” said OpenAI CEO Sam Altman in a post on X. “Available today in ChatGPT and the API. It is the smartest generally-available model in the world, and in particular is good at doing real-world knowledge work tasks.”
“Already, the average ChatGPT Enterprise user says AI saves them 40–60 minutes a day, and heavy users say it saves them more than 10 hours a week,” OpenAI said in a blog post. “We designed GPT‑5.2 to unlock even more economic value for people; it’s better at creating spreadsheets, building presentations, writing code, perceiving images, understanding long contexts, using tools, and handling complex, multi-step projects.”
GPT-5.2 Thinking significantly surpasses GPT-5.1 Thinking in a wide range of benchmarks, particularly in knowledge work tasks and abstract reasoning. It also demonstrated improvements in software engineering, math and science problems.
“GPT-5.2 represents the biggest leap for GPT models in agentic coding since GPT-5 and is a SOTA coding model in its price range,” said Windsurf CEO Jeff Wang. “The version bump undersells the jump in intelligence.”
GPT‑5.2 is priced at $1.75/1M input tokens and $14/1M output tokens, with a 90% discount on cached inputs. OpenAI said that despite GPT‑5.2’s higher cost per token, the cost of attaining a given level of quality ended up less expensive due to GPT‑5.2’s greater token efficiency.
GPT-5.2 was built in collaboration with OpenAI’s partners Microsoft (MSFT) and Nvidia (NVDA) utilizing Azure data centers powered by a range of Nvidia’s GPUs, including the H100, H200 and GB200-NVL72.
“Super excited about GPT-5.2 from our partners at OpenAI,” said Microsoft CEO Satya Nadella in a post on X. “Making it natively part of the Microsoft tools people use every day along with your work data helps all the new model capabilities shine. The pace of progress continues to accelerate.”
The release of GPT-5.2 follows recent launches by rivals Anthropic (ANTHRO) and Google (GOOG)(GOOGL), which unveiled Claude Opus 4.5 and Gemini 3, respectively, in November.