OpenAI has integrated GPT-4.1 and GPT-4.1 mini into ChatGPT, with the latter replacing GPT-4o mini for all users.
GPT-4.1 is a specialised model focused on programming tasks and instruction execution. It boasts enhanced speed and serves as an alternative to OpenAI o3 for developers’ everyday needs.
Access to the model has been granted to subscribers of the Plus, Pro, and Team plans. It will become available for Enterprise and Edu users in the coming weeks.
In April, OpenAI introduced a new family of AI models geared towards programming and instruction execution—GPT-4.1, GPT-4.1 mini, and GPT-4.1 nano.
GPT-4.1 outperforms GPT-4o and GPT-4o mini in programming benchmarks, including SWE-bench. GPT-4.1 mini and nano are more efficient and faster, albeit with a slight reduction in accuracy. Initially, they were only accessible via API.
The company faced criticism from researchers for releasing GPT-4.1 without a safety report. Critics claimed the startup was lowering its transparency standards. In response, the firm noted that the models are not cutting-edge.
3/ While this is a notable improvement, GPT-4.1 doesn’t introduce new modalities or ways of interacting with the model, and doesn’t surpass o3 in intelligence. This means that the safety considerations here, while substantial, are different from frontier models.
— Johannes Heidecke (@JoHeidecke) May 14, 2025
“GPT-4.1 doesn’t introduce new modalities or ways of interacting with the model and doesn’t surpass o3 in intelligence. This means that the safety considerations here, while substantial, are different from frontier models,” commented Johannes Heidecke, head of OpenAI’s safety systems department.
Earlier that month, OpenAI released the deceptive o3 and o4-mini.