OpenAI has officially launched its GPT-4.1 and GPT-4.1 Mini models within ChatGPT, expanding access to its latest AI advancements for developers and everyday users alike.
According to OpenAI, GPT-4.1 delivers superior performance in coding and instruction-following compared to previous models, including GPT-4o. It’s also significantly faster, making it an ideal tool for software engineers who rely on ChatGPT for writing, debugging, and testing code.
The GPT-4.1 model is now available to users subscribed to ChatGPT Plus, Pro, and Team plans. In parallel, GPT-4.1 Mini—designed for speed and lightweight performance—is rolling out to both free and paid users. This update replaces GPT-4.0 Mini, which is being phased out across all ChatGPT accounts.
While OpenAI initially launched these models through its developer API in April, the release was met with some pushback from the AI research community for lacking an accompanying safety report. OpenAI defended the move, stating that GPT-4.1, while powerful, doesn’t introduce new interaction methods or exceed the capabilities of more advanced frontier models like GPT-4o3.
Johannes Heidecke, OpenAI’s Head of Safety Systems, clarified that while safety remains a priority, the level of risk presented by GPT-4.1 is substantially different from that of cutting-edge frontier systems.
In response to ongoing transparency concerns, OpenAI also announced the launch of its Safety Evaluations Hub, a new platform where the company will regularly publish results from its internal model safety assessments. This initiative marks a step toward greater openness in AI development and deployment.
The timing of the rollout is notable as competition intensifies in the AI coding space. OpenAI is rumored to be nearing a $3 billion acquisition of Windsurf, a leading AI tool for developers. Meanwhile, Google has upgraded its Gemini chatbot to integrate more seamlessly with GitHub.
With these advancements, OpenAI continues to push forward in making generative AI more useful, transparent, and accessible to both technical users and the broader public.
Source: (Techcrunch)
