OpenAI Unveils GPT-5: A Unified AI Model Merging Reasoning & Multimodality, Launching This Summer

OpenAI is preparing to unify its models and unveil GPT-5 later this summer. The company has officially confirmed that its next step will not merely be the release of a new large-scale language model, but a comprehensive integration of its two primary product lines — GPT and the “o”-series.

Currently, multiple models operate within ChatGPT, each tailored to distinct tasks. Despite their impressive capabilities, users frequently encounter confusion, as the models bear similar names and their individual purposes are not always transparent.

Of particular note is OpenAI’s “o”-series, which has been designed to tackle complex reasoning tasks. Models such as GPT-4o and others in this line are optimized for multimodal interactions — they can process not only text, but also images, audio, and video.

In the forthcoming GPT-5, the developers aim to synthesize the strengths of both approaches. As Romain Huet, OpenAI’s Head of Developer Experience, emphasized, the company is not merely building another groundbreaking AI — its ultimate goal is to consolidate the advances of both series into a single, cohesive model. According to him, GPT-5 will marry the reasoning prowess of the “o”-series with the multimodal versatility of the GPT line.

OpenAI spokesperson Jerry Tworek previously remarked that GPT-5 is intended not just as the company’s next foundational model, but as a unified solution capable of performing all current model functions more effectively and intuitively — eliminating the need to switch between different versions of the AI. This unification is expected to significantly streamline user experience and enhance overall clarity.

While an exact release date has not yet been announced, OpenAI CEO Sam Altman has already hinted that GPT-5 will debut sometime this summer.