OpenAI has officially confirmed that it is working on GPT-5 — a new universal model that will combine breakthroughs in multimodality and logical thinking inherent in different lines of current models. The release is expected to take place this summer. The goal is to create a single system that can perform all tasks without switching between models.
In the current ChatGPT, users deal with several models: some are optimized for logic (the “o” series), others for multimodality (GPT-4o). This creates confusion and the need to switch depending on the task.
In GPT-5, OpenAI promises to combine these lines into one new “frontier” model that will combine:
understanding complex logic and chains of reasoning;
processing text, images, audio, and possibly video;
reducing the need to switch models;
a unified interface for all types of tasks.
Company representatives noted:
> “We’re not just creating a new model — we’re unifying our two series,” said Romain Huet, head of OpenAI’s development department.
Vice President Jerry Tworek also confirmed on Reddit that GPT-5 is a fundamental update that should make all current capabilities better, simpler, and more efficient.
In parallel, OpenAI is testing other features — including a Study Together mode for ChatGPT, which will allow students to study together or use the AI as a partner for preparation.
With GPT-5, OpenAI is taking an important step towards unified artificial intelligence that can handle all types of tasks — from analytics to creativity — within a single model. This is not just a technical upgrade — it’s an attempt to get closer to true AGI (Artificial General Intelligence). If all goes according to plan, this summer users will receive a unified tool that not only combines functions, but also rethinks the principle of working with AI.
