OpenAI has announced the release of its newer GPT-4.1 model, which is now available to all paying ChatGPT users. Starting today, subscribers on ChatGPT Plus, Pro and Team plans will gain access to it, which can be selected via the “more models” dropdown menu. Enterprise and Edu subscribers, however, will have to wait a few more weeks before the update reaches them.
GPT-4.1 is touted to offer improvements in speed and efficiency. In addition to its release, a scaled-down variant called GPT-4.1 mini will be replacing the current GPT-4o mini. OpenAI says this will apply across all user tiers, including those on the free plan.
By popular request, GPT-4.1 will be available directly in ChatGPT starting today.
GPT-4.1 is a specialized model that excels at coding tasks & instruction following. Because it’s faster, it’s a great alternative to OpenAI o3 & o4-mini for everyday coding needs.
— OpenAI (@OpenAI) May 14, 2025
Once free-tier users hit the usage limit for the more powerful GPT-4o, GPT-4.1 mini will serve as their default fallback model. However, the company did not say when this change is expected to rollout.
On a related note, OpenAI has also introduced a new “safety evaluations hub,” a public webpage designed to increase transparency around the behaviour of its AI systems. The hub includes data on model performance in areas such as hallucination frequency, compliance with instructions, production of harmful content, and resistance to jailbreak attempts.
Introducing the Safety Evaluations Hub—a resource to explore safety results for our models.
While system cards share safety metrics at launch, the Hub will be updated periodically as part of our efforts to communicate proactively about safety.https://t.co/c8NgmXlC2Y
— OpenAI (@OpenAI) May 14, 2025
“As the science of AI evaluation evolves, we aim to share our progress on developing more scalable ways to measure model capability and safety,” OpenAI explained. The company hopes the new hub will not only provide insight into the current safety profile of its models — such as GPT-4.1 and GPT-4.5 — but also support broader transparency efforts within the AI research community.
OpenAI emphasised that the safety hub offers only a partial view of model performance, describing it as a “snapshot.” For a more complete understanding, it recommends consulting additional resources such as system cards, technical assessments, and official documentation.
