OpenAI unveils new GPT-4.1 – faster, smarter, cheaper

OpenAI has announced an updated flagship artificial intelligence model, GPT-4.1. This release is a follow-up to last year’s GPT-4o and marks an important step in the development of multi-format AI systems, especially against the backdrop of the delayed launch of GPT-5.

OpenAI representatives emphasized during the live broadcast that GPT-4.1 outperforms GPT-4o in virtually every aspect, from speed and accuracy to task versatility.

Three models in one lineup: Mini to Nano

Along with the basic GPT-4.1 version, the company introduced two lighter versions aimed at developers and embedded solutions:

• GPT-4.1 Mini is a smaller and more affordable version that, like its predecessor, is aimed at a broad audience of developers looking for a balance between cost and functionality
• GPT-4.1 Nano is the lightest and fastest model released by OpenAI. The developers call it “our smallest, fastest, and cheapest AI.”

The two models emphasize cost-effectiveness and speed, opening up new opportunities for embedding AI in mobile and cloud solutions.

GPT-4 becomes a thing of the past, GPT-4o becomes the foundation

OpenAI also confirmed that the original GPT-4, launched two years ago, will be retired from ChatGPT as early as April 30. It will be replaced by GPT-4o, which recently received a major upgrade – including image generation features that caused such a stir that the company had to limit the number of requests and suspend access to free accounts to avoid overloading the GPU.

With these improvements, GPT-4o is the “natural successor” to GPT-4, as noted in the official OpenAI changelog.

GPT-5 is delayed – but there’s more to come

Although there has been much discussion recently about the possible imminent launch of GPT-5, OpenAI CEO Sam Altman told social network X on April 4 that the release has been delayed “for several months.” He said the company has encountered unexpected difficulties in integrating all of the system’s components.

At the time, he said, the company has been struggling with the integration of all of the system’s components.