OpenAI has announced that, following popular demand, the GPT-4.1 model is now available for use directly within ChatGPT, starting today.
GPT-4.1 is highlighted as a specialized model with particular strengths in handling coding tasks and following instructions effectively. The company notes that due to its enhanced speed, GPT-4.1 presents a strong alternative to OpenAI’s o3 and o4-mini models for users engaging in daily coding activities.
OpenAI announced three new models GPT‑4.1, GPT‑4.1 mini, and GPT‑4.1 nano in the form of API last month. These models outperform GPT‑4o and GPT‑4o mini across the board, with major gains in coding and instruction following, said the company. Further, these offer good performance at low latency, have larger context windows, and a refreshed knowledge cutoff of June 2024.
GPT‑4.1, GPT‑4.1 mini, and GPT‑4.1 nano: Coding Enhancements
GPT‑4.1 is significantly better than GPT‑4o at a variety of coding tasks, including agentically solving coding tasks, frontend coding, making fewer extraneous edits, following diff formats reliably, ensuring consistent tool usage, and more.
When it comes to benchmarks, OpenAI says, GPT‑4.1 scores 54.6% on SWE-bench Verified, improving by 21.4% over GPT‑4o and 26.6% over GPT‑4.5, making it a leading model for coding.
GPT‑4.1 follows diff formats more reliably, which allows developers to save both cost and latency by only having the model output changed lines, rather than rewriting an entire file. Further, these new models excel at following formats properly, which reduces frequent edits.
Long Context Awareness
Another main highlight of these new models is their long context awareness, which the company feels is very important for developers who work on apps across legal, coding, customer support, and many other domains.
GPT‑4.1, GPT‑4.1 mini, and GPT‑4.1 nano can process up to 1 million tokens of context—up from 128,000 for previous GPT‑4o models. Thanks to increased tokens, GPT‑4.1 is far more reliable at noticing relevant text and ignoring distractors across long and short context lengths.
To illustrate this with a benchmark, on Video-MME (opens in a new window)—a benchmark for multimodal long-context understanding—GPT-4.1 sets a new state-of-the-art result, scoring 72.0% in the “long, no subtitles” category, a 6.7% absolute improvement over GPT-4o.
Following Instructions and Low Latency
After collecting feedback from developers, OpenAI has developed an internal eval for instruction following to track model performance across several dimensions like format following, content requirements, ranking, overconfidence, and more.
On Scale’s MultiChallenge benchmark, a measure of instruction following ability, GPT‑4.1 scores 38.3%, a 10.5% increase over GPT‑4o.
When it comes to performance, GPT 4.1 models offer great performance at lower latency. GPT‑4.1 mini, for instance, matches or exceeds GPT‑4o in intelligence evals while reducing latency by nearly half and reducing cost by 83%.
For tasks that demand low latency, GPT‑4.1 nano is the company’s fastest and cheapest model available now. It delivers exceptional performance at a small size with its 1 million token context window, making it ideal for tasks like classification or autocompletion.
These improvements in instruction following, reliability, low latency, and long context comprehension make GPT‑4.1 models considerably more effective at powering agents, or systems that can independently accomplish tasks on behalf of users.
GPT-4.5 Preview API to be discontinued
Since GPT‑4.1 offers improved or comparable performance to GPT-4.5 Preview on many key capabilities at a much lower cost and latency, GPT‑4.5 Preview will be turned off on July 14, 2025.
Furthermore, OpenAI states that in ChatGPT, many of the improvements in instruction following, coding, and intelligence have already been gradually incorporated into the latest version of GPT‑4o, and more will continue to be added in future releases.
Availability
GPT‑4.1, GPT‑4.1 mini, and GPT‑4.1 nano are available now to all directly in ChatGPT.
GPT-4.1 mini, replaces GPT-4o mini, in ChatGPT for all users. Plus, Pro, and Team users can access GPT-4.1 via the “more models” dropdown in the model picker.
OpenAI said that Enterprise and Edu users will get access in the coming weeks. The company said that the GPT-4.1 has the same rate limits as GPT-4o for paid users.