OpenAI Launches GPT-5.2 With Instant, Thinking & Pro Modes for Faster, More Accurate Professional Work

OpenAI has released GPT-5.2 for ChatGPT users, offering three new modes aimed at professional work. The updated model, described by OpenAI as its strongest yet for everyday use, is now rolling out. Paid subscribers gain access to GPT-5.2 Instant, GPT-5.2 Thinking and GPT-5.2 Pro within the ChatGPT interface.

The three GPT-5.2 versions began reaching paying customers on 12 December 2025. OpenAI says these options are meant to cover different needs, from quick chats to complex projects. All variants sit inside ChatGPT, which users access through the familiar app and web layouts on phones and computers.

GPT-5.2 in ChatGPT: key modes and features

Among the new options in ChatGPT, OpenAI highlights GPT-5.2 Thinking as the standout. On the GDPval benchmark, covering professional-level tasks across 44 occupations, this mode sets a new score record. Experts report it matches or exceeds human specialists in roughly 71 per cent of comparisons.

GPT-5.2 Instant is described by OpenAI as the rapid, everyday model for ChatGPT. It aims to deliver faster replies with clearer explanations than before. OpenAI says it performs better on how-to questions, technical writing and translation tasks, giving more direct and readable answers for routine queries.

GPT-5.2 Thinking and Pro in ChatGPT for heavy work

GPT-5.2 Thinking targets demanding workloads across ChatGPT. It is built for coding, lengthy document summaries, maths, logic and detailed planning, where structure matters. OpenAI says this mode generates richer, more thorough responses. Early tests show it also improves at front-end development and complicated user interface designs, including 3D components.

GPT-5.2 Pro sits at the top of the ChatGPT range for users who prioritise accuracy over response speed. According to OpenAI, trials suggest this model makes fewer serious mistakes, especially on challenging programming questions. It is aimed at people who rely on ChatGPT for careful analysis, complex systems work and other high-risk tasks.

Across Instant, Thinking and Pro, GPT-5.2 improves reasoning, long-context understanding, tool use and image handling inside ChatGPT. OpenAI reports around 30 per cent fewer incorrect answers compared with GPT-5.1 Thinking. The company says this boosts reliability for research, writing, analytics and decision support, where consistency and fewer errors are important for professionals.

Vision-related skills in GPT-5.2 Thinking also see major changes. OpenAI says this version roughly halves error rates when reading charts, dashboards, diagrams and software interfaces. That could aid finance, engineering, operations, design and customer support teams, which often rely on visuals. The gains apply when ChatGPT interprets screenshots, product flows or monitoring panels.

New spreadsheet and presentation tools in ChatGPT work only with GPT-5.2 Thinking or GPT-5.2 Pro. Users need a Plus, Pro, Business or Enterprise subscription to try these features. OpenAI warns that generating complex files might take several minutes, especially when ChatGPT handles large datasets or detailed presentation layouts from uploaded documents.

The GPT-5.2 launch comes as OpenAI and Sam Altman face pressure from rivals. Earlier reports described how competition led Altman to declare a “code red”. Another report said OpenAI had also labelled ChatGPT “code red” as challengers grew stronger. Alongside this, OpenAI continues training models while global news events dominate headlines.

For Indian users following technology, GPT-5.2 marks another step in how ChatGPT is used for work. With three modes tuned for speed, depth or precision, OpenAI aims to support tasks from translations to advanced programming. The company’s benchmark results, tool upgrades and visual improvements suggest a focus on reliability across many professional settings.