Date:
April 15, 2025
OpenAI has launched three new AI models in its API: GPT-4.1, GPT-4.1 mini, and GPT-4.1 nano. OpenAI’s new GPT-4.1 AI models focus on coding tasks efficiently.
On Apr 14, 2025, OpenAI has published an official document announcing its new AI models for developers, which are much advanced compared to GPT-4o and GPT-4o mini. These new models are also capable of offering a better context that will support up to 1 million tokens and have a better comprehension ability.
Here are the industry standards excelled by GPT-4.1 compared to other models.
| Category | Benchmark | GPT‑4o | GPT‑4.5 | GPT‑4.1 | Improvement (vs GPT‑4o) | Improvement (vs GPT‑4.5) |
| Coding | SWE-bench Verified | 33.20% | 28.00% | 54.60% | 0.214 | 0.266 |
| Instruction Following | Scale’s MultiChallenge | 27.80% | – | 38.30% | 0.105 | – |
| Long Context (Multimodal) | Video-MME (long, no subtitles) | 65.30% | – | 72.00% | 0.067 | – |
OpenAI has partnered with developers to understand the real-world utilities to build GPT 4.1. The firm claims that this new model family is supposed by cost-efficient, yet much better in performance compared to predecessors like 4o.
OpenAI states in its official GPT-4.1 release document that:
“GPT‑4.1 mini is a significant leap in small model performance, even beating GPT‑4o in many benchmarks. It matches or exceeds GPT‑4o in intelligence evals while reducing latency by nearly half and reducing cost by 83%.”
Open AI has also said describing the nature of its benchmarks:
“While benchmarks provide valuable insights, we trained these models with a focus on real-world utility. Close collaboration and partnership with the developer community enabled us to optimize these models for the tasks that matter most to their applications.”
GPT-4.1 mini is competitively better than the quality of even GPT-4o. The 4.1 model is great for tasks requiring low latency, scoring 80.1% on MMLU, 50.3% on GPQA, and 9.8% on Aider polyglot coding.
These scores are higher compared to GPT-4o mini and great for classification tasks or even autocompletion.
As for the availability of GPT 4.1, OpenAI has clearly stated,
“Note that GPT‑4.1 will only be available via the API. In ChatGPT, many of the improvements in instruction following, coding, and intelligence have been gradually incorporated into the latest version(opens in a new window) of GPT‑4o, and we will continue to incorporate more with future releases.”
