OpenAI has launched its new GPT-4.1 and GPT-4.1 mini AI models, now available in ChatGPT. The release, announced on Wednesday, aims to enhance the experience for software engineers using ChatGPT for tasks such as coding and debugging. According to OpenAI, the GPT-4.1 models are designed to outperform the GPT-4o series in coding and instruction-following, offering increased speed and improved performance for developers.
The models are now being rolled out to subscribers of ChatGPT Plus, Pro, and Team, while GPT-4.1 mini will be available for free and paid users of the platform. With this update, OpenAI has removed the previous GPT-4.0 mini version. GPT-4.1 and GPT-4.1 mini were initially released in April, but only through OpenAI’s developer API. The models were met with some criticism from the AI research community at the time, as there was no safety report released alongside them, with concerns raised about the transparency of OpenAI’s approach. However, OpenAI clarified that GPT-4.1 was not considered a frontier model and therefore did not require the same level of safety reporting.
You May Also Like: OpenAI Reverses Update That Made ChatGPT “Overly Sycophantic”
To improve transparency, OpenAI also announced that it will publish internal safety evaluations of its AI models more frequently. These evaluations will be accessible through a new Safety Evaluations Hub launched by OpenAI on Wednesday. This move comes amid growing attention on AI tools in the coding space, with OpenAI’s ongoing efforts to integrate AI into the development process, including its reported $3 billion acquisition of AI coding tool Windsurf.
This update to ChatGPT highlights OpenAI’s continuing focus on improving its AI’s capabilities in software development, particularly in enhancing speed and accuracy for coding tasks. By making the GPT-4.1 models more accessible, OpenAI aims to support a broader community of developers while also addressing transparency concerns raised by the AI research community.
