OpenAI's New GPT-5.2 Model Wants To Help You Automate Your Job

OpenAI has released GPT-5.2, which aims to be ChatGPT’s “most capable model series yet for professional knowledge work.”

It comes a few weeks after the launch of GPT-5.1, and seemingly drops the “garlic” codename that OpenAI teased online this week. Examples of what it can do include coding, creating spreadsheets, building presentations, and handling complex, multi-step projects. OpenAI is leaning into its workplace customers, and claims ChatGPT enterprise users can save 40–60 minutes a day, or more than 10 hours a week for heavy users, with GPT-5.2.

While ChatGPT can already do all these things in some form, OpenAI says GPT-5.2 “sets a new state of the art across many benchmarks,” especially the 44 occupations for which it tested its performance on “well-specified” tasks. Its real-world performance remains to be seen. On X, CEO Sam Altman admitted it can’t do everything, including “output polished files.”

This Tweet is currently unavailable. It might be loading or has been removed.

GPT-5.2 is rolling out today to paid plans, with three variants: Instant, Thinking, and Pro. The Thinking model “beats or ties top industry professionals on 70.9% of…knowledge work tasks, according to expert human judges,” OpenAI says. In internal testing, it did the tasks at more than “11x the speed” and less than “1% the cost of expert professionals, suggesting that when paired with human oversight, GPT‑5.2 can help with professional work.” However, speed and costs estimates may vary.

A spreadsheet made by GPT-5.1 Thinking and a more complex version by GPT-5.2 Thinking (Credit: OpenAI)

A holiday card coded by GPT-5.2

A holiday card coded by GPT-5.2 (Credit: OpenAI)

Last week, OpenAI declared an internal “code red” after Google’s latest Gemini 3 model set new industry benchmarks for performance. OpenAI has since directed more company resources to improve ChatGPT, but says it did not speed up the launch of GPT-5.2 because of competition from Google, according to Wired. However, its CEO of applications, Fidji Simo, noted the additional resources for ChatGPT have been “helpful.”

In fairness, GPT-5.2’s focus on business and workplace users does not seem to be a direct competitor to Gemini 3, which has a reputation as a general-purpose chatbot. Companies like Shopify and Zoom have already been testing GPT-5.2, according to OpenAI.

Recommended by Our Editors

Altman says Gemini 3 had less of an effect on the company’s metrics than expected, and that the company should exit the code red by January, CNBC reports.

Disclosure: Ziff Davis, PCMag’s parent company, filed a lawsuit against OpenAI in April 2025, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.