OpenAI launches GPT-5.2 AI model to return fire on Google Gemini 3

OpenAI unveiled its latest artificial intelligence (AI) model, GPT-5.2, on Thursday after a 'red alert' on Google Gemini 3 eating into its business.

Available within OpenAI's ChatGPT chatbot, the latest model can create spreadsheets and presentations, decode images, write code, and even understand long context, the company said.

GPT-5.2 is available in three versions — Instant, Thinking, and Pro — starting with the paid plans. In the API, they are available now to all developers.


Each version will have specific capabilities. Instant will deliver quick responses for writing and information extraction. Thinking will take over coding- and planning-related tasks, whereas the Pro version will handle complex queries, the Sam Altman-led company said.

Coding chops

On coding tasks — which have fast become a popular enterprise revenue stream for AI companies — OpenAI claims GPT-5.2 can more reliably debug production code, implement feature requests, refactor large codebases, and ship fixes end-to-end with less manual intervention.

The company said GPT‑5.2 Thinking performs better at front-end software engineering than its predecessors, as early testers found it more capable at front-end development and complex UI work, especially involving 3D elements.

'Red alert' response

OpenAI had launched GPT 5.1 model on November 12, with Google following with its Gemini 3 less than a week later. In a note to employees, Sam Altman, the chief executive of the ChatGPT maker, had warned employees of "economic headwinds" due to the latest surge in competition.

He did assure that the company is catching up fast. “We have built enough strength as a company to weather great models shipping elsewhere competition... (so), having most of our research team focussed on really getting to superintelligence is critically important,” he mentioned.