In a significant move for artificial intelligence enthusiasts, OpenAI has officially brought its much-anticipated GPT-4.1 models to ChatGPT. Announced today, May 14, 2025, the rollout gives paid subscribers immediate access to the company’s latest AI advancements.

New Models for Paid and Free Users

Starting today, all Plus, Pro, and Team subscribers can access GPT-4.1 through the “more models” dropdown in ChatGPT’s interface. Enterprise and Edu users won’t be left out for long, with OpenAI promising access “in the coming weeks.” The model maintains the same rate limits as GPT-4o for all paid users.

Alongside the flagship model, OpenAI has introduced GPT-4.1 mini, which replaces GPT-4o mini. This smaller, more efficient model is available to paid users and serves as the fallback option for free users who exhaust their GPT-4o usage limits.

“GPT-4.1 mini is a fast, capable, and efficient small model, delivering significant improvements compared to GPT-4o mini—in instruction-following, coding, and overall intelligence,” according to OpenAI’s release notes.

Technical Improvements

The GPT-4.1 family represents a substantial leap forward in AI capabilities. Since its API launch in April, developers have embraced the model for its enhanced performance, particularly in coding tasks.

The technical improvements are substantial. On the SWE-bench Verified benchmark, which evaluates real-world software engineering abilities, GPT-4.1 achieves 54.6% accuracy—a dramatic 21-point increase over GPT-4o’s 33.2%, as InfoQ reported.

The model also excels at following precise instructions and handling complex web development tasks, making it an attractive alternative to specialized coding models. OpenAI’s internal evaluations demonstrate GPT-4.1’s superior ability to process extended inputs, with all models in the family capable of handling up to 1 million tokens.

New Transparency in Safety Evaluations

Coinciding with this release, OpenAI has launched a new Safety Evaluations Hub, enhancing transparency around model capabilities and limitations. This resource provides detailed results from the company’s standard safety evaluations, giving users greater insight into how models like GPT-4.1 perform across various metrics.

“We believe providing more detail about our safety evaluations is an important step toward building trust with users and the broader AI community,” the company stated in its announcement.

Pushing the Competition

The release comes at a time of intense competition in the AI space. Experts anticipate GPT-4.1 will further solidify OpenAI’s leadership position while pushing competitors to accelerate their own development timelines.

Early user reactions have been predominantly positive. Johannes Heidecke, an AI researcher, shared on social media that the model shows “impressive improvements in coding” and praised its enhanced reasoning capabilities, especially for complex tasks.

OpenAI’s Dual Strategy

This rollout marks another milestone in the rapid evolution of consumer AI tools. The introduction of both premium and more accessible models reflects OpenAI’s dual strategy of pushing the boundaries of what’s possible while ensuring broader access to AI capabilities.

For businesses and developers already using OpenAI’s technologies, the upgrade path is clear. The enhanced capabilities of GPT-4.1, particularly in coding and instruction following, open new possibilities for automation, content creation, and problem-solving applications.

As AI tools continue to advance at a breathtaking pace, today’s release reminds us that we’re still in the early chapters of the artificial intelligence revolution. With each iteration, these systems become more capable, more accessible, and increasingly integrated into our digital lives.