OpenAI has unveiled its new flagship artificial intelligence model, GPT-5.2, aimed at professional users and the corporate sector. The company emphasizes that this version is the most advanced in its lineup and offers significantly enhanced capabilities for data handling, programming, analytics, and business process automation.
This is reported by Finway
Features and Functionality of GPT-5.2
GPT-5.2 is available for payment both through an API for developers and to ChatGPT Plus users. The model is presented in three variants for different use cases:
- Instant — for performing quick tasks such as information retrieval or text generation;
- Thinking — focused on complex analytical operations, programming, document handling, and mathematical calculations;
- Pro — provides maximum accuracy and reliability for mission-critical corporate tasks.
According to OpenAI’s Chief Product Officer, Fiji Simo,
“GPT-5.2 was designed with practical utility in mind: it generates tables, presentations, code more efficiently, analyzes images, works with large contexts, and helps manage multi-layered projects.”
Competition with Google and Cybersecurity Risks
The launch of GPT-5.2 comes amid heightened competition between OpenAI and Google, whose product Gemini 3 Pro maintains leadership in key rankings, except in the programming sector. It is known that recently OpenAI’s CEO Sam Altman announced a “red code” mode to accelerate the improvement of AI models due to declining ChatGPT traffic and the risk of losing market positions.
Some analysts report that the company considered delaying the release of GPT-5.2 for additional testing; however, OpenAI ultimately chose a rapid launch strategy to strengthen its position in the corporate segment.
According to internal assessments by OpenAI, GPT-5.2 shows improved results in programming, mathematics, scientific tasks, computer vision, and handling large volumes of context. The Thinking mode, according to the company’s tests, outperforms Gemini 3 Pro and Claude Opus 4.5 in tasks requiring logical reasoning, software development, and scientific benchmarks.
Experts note that the focus on improving reasoning in the GPT-5.2 models has led to increased computational costs. The company, according to media reports, is already spending more on inference than previously reported, paying for computational resources directly rather than through subsidized cloud services. OpenAI expects to offset these costs by scaling products and increasing revenues, while the efficiency of the models continues to grow.
OpenAI also warns that the capabilities of their new AI models are developing faster than developers had anticipated, and future generations may pose a “high” risk in the field of cybersecurity.