With GPT-5, OpenAI aims to end the fragmentation of AI functionalities and models within ChatGPT.
OpenAI, the company behind ChatGPT, is preparing to launch GPT-5, the next generation of its generative AI model, which underpinned the first ChatGPT version. Since then, the AI models have increasingly diverged.
For example, ChatGPT currently uses different models for different tasks, such as text generation (GPT-4 and GPT-4.5), image creation (DALL-E), speech recognition (Whisper), programming tasks (Codex), text-to-video (Sora), and code analysis. OpenAI wants to bundle all these functionalities into a single, integrated system.
Jerry Tworek, vice president at Openai, writes in a post on Reddit that GPT-5 is set to become the ” next foundational model. ” It should primarily make things easier for the user. They will no longer have to choose a model based on the task at hand: GPT-5 will determine which underlying model is used to complete each task successfully. All of this is in the same interface.
If OpenAI follows through with that plan, it would fit into a broader trend in the AI industry. Companies are pursuing so-called “agentic AI”: systems that can reason, plan, and make decisions autonomously about which tools to use to perform a task. GPT-5 could be one such step toward a more self-driving AI assistant.
OpenAI has yet to announce an official launch date for GPT-5, but insiders expect it to roll out later this year.