OpenAI Introduces GPT-4.1 Models in ChatGPT for Developers

OpenAI has launched its GPT-4.1 and GPT-4.1 mini models within ChatGPT, the company announced on X. These models aim to improve the coding and debugging capabilities for software engineers using the platform. According to OpenAI, GPT-4.1 outperforms GPT-4.0 in coding and instruction following, delivering faster results compared to the previous version. The new GPT-4.1 models are being rolled out to subscribers of ChatGPT Plus, Pro, and Team, while GPT-4.1 mini is available for both free and paid users. Following this update, GPT-4.0 mini will be removed from the platform. OpenAI initially introduced the models in April through its API for developers, drawing criticism from the AI research community for not releasing a safety report. Critics argued that OpenAI was compromising transparency by shipping GPT-4.1 without adequate safety documentation. However, OpenAI defended the decision, stating that GPT-4.1, while faster and more efficient than GPT-4.0, did not reach the level of "frontier models" and thus did not require the same safety reports. In response to growing concerns, OpenAI has committed to publishing internal AI model safety evaluations more frequently. The results will be available in the newly launched Safety Evaluations Hub. This release coincides with increased interest in AI coding tools, with reports suggesting OpenAI is close to acquiring Windsurf, a leading AI coding tool, for $3 billion. Additionally, Google has updated its Gemini chatbot to integrate more seamlessly with GitHub projects, signaling further developments in AI-driven coding tools.