OpenAI has introduced its latest artificial intelligence model, GPT-4.1, along with its lighter versions, GPT-4.1 mini and GPT-4.1 nano. These models are smarter, more powerful, and more cost-efficient than previous versions.
According to OpenAI, the new models show major improvements in coding, instruction following, and understanding long texts. The GPT-4.1 model has even outperformed the company’s previous most powerful model, GPT-4o. It can process up to 1 million tokens at once — tokens are the units of text that AI models work with. This means GPT-4.1 can better analyze and respond to very long documents or complex questions.
The model's knowledge base has also been updated through June 2024, allowing it to provide more relevant and up-to-date answers.
OpenAI CEO Sam Altman wrote on social media:
“The model performed very well in tests, but our main goal was to make it useful in real-world use. Developers are very happy with the new models.”
The company stated that GPT-4.1 is more affordable than previous models while offering better performance. For this reason, the GPT-4.5 API version will be discontinued in July.
OpenAI initially released GPT-4.5 as a research preview in February for selected users. Now, GPT-4.1 will fully replace it.
