OpenAI has unveiled a new GPT-5.2 model, which the company calls the most powerful and accurate in history
GPT-5.2 is already available to ChatGPT users and developers via the API.
GPT-5.2 is already available to ChatGPT users and developers via the API.
Almost immediately after the announcement of the “code red” , OpenAI announced the launch of GPT-5.2 — an updated large language model that significantly improves the quality of work with text, code, large documents, and multimodal tasks.
The company claims that this is the most capable model for professional tasks, setting new records in a number of tests and benchmarks. In particular, GPT-5.2 demonstrated maximum results on the AIME-2025 math test and significantly outperformed previous generations in logic and analysis tasks.
In the new version, OpenAI has focused more on the model’s ability to work with long contexts and tools from generating spreadsheets and presentations to complex software solutions. The company also introduced three use cases: Instant (quick answers), Thinking (deep analytics), and Pro (maximum quality for professional users).
This is a key release for OpenAI amid increasing competition in the AI market. The improved performance of GPT-5.2 should strengthen the company's position compared to the Gemini model, which Google has integrated into its new AI browser. The model was also created in partnership with Nvidia and Microsoft, which provided high-performance infrastructure.
As a reminder, Disney has signed a three-year licensing agreement with OpenAI for over 200 characters from the Marvel, Star Wars, and Pixar universes, which will officially appear in the image and video generation tools in Sora and ChatGPT.
OpenAI and Anthropic launch Agentic AI Foundation to advance agent AI interoperability standards
OpenAI and Anthropic launch Agentic AI Foundation to advance agent AI interoperability standards
OpenAI's ChatGPT Lead Says Posts That Allege In-App Ads Are "Not Real or Ads"
OpenAI's ChatGPT Lead Says Posts That Allege In-App Ads Are "Not Real or Ads"
OpenAI's leadership in the AI market has been shaken: Sam Altman declared a "code red"
OpenAI's leadership in the AI market has been shaken: Sam Altman declared a "code red"
Read the country's main IT news in our Telegram
Read the country's main IT news in our Telegram
