OpenAI introduces new-gen GPT models, here’s what makes them special

Sam Altman led AI Startup, OpenAI, has introduced new generation AI models, the GPT‑ 4.1, GPT‑ 4.1 mini, and GPT‑ 4.1 nano. The GPT‑ 4.1 takes the flagship AI model crown and is the successor to the GPT- 4o model with powerful capabilities and a larger context window. All the new AI models are capable of understanding text, images, or videos, all providing up to 1 million context length, which means it can take about 750,000 words in one go, much more powerful than GPT- 4o’s 128,000-token limit. During the launch, OpenAI highlighted that the new generation AI models focus on coding and instruction following. Therefore, know in detail about OpenAI’s new AI models.

Also read: 7 Ways AI is Transforming Creative Industries

OpenAI launches GPT-4.1 AI models

OpenAI launched the flagship GPT- 4.1 AI model along with two smaller models, the GPT‑ 4.1 mini, and GPT‑ 4.1 nano. The company highlights that the new GPT- 4.1 model is more capable of understanding complex prompts in comparison to the GPT- 4.0 model, making it more reliable. The new model is more capable of managing front-end coding, allowing developers to create web apps which are “more functional and aesthetically pleasing.” In a real-world software engineering skills test, the GPT‑ 4.1 completed over 54.6% of tasks, whereas the GPT- 4o completed 33.2% of tasks, showcasing a significant improvement.

Also read: 5 Google Pixel AI features that will make you rethink your smartphone experience right now

Coming to the other two models, the GPT‑ 4.1 mini and GPT- 4.1 are efficient and more cost-effective AI models, providing improved performance than the flagship GPT-4o model in several tasks. Furthermore, OpenAI claims that GPT- 4.1 nano is the fastest and cheapest AI model. It was revealed that the AI model achieved 80.1% on MMLU, 50.3% on GPQA, and 9.8% on Aider polyglot coding. OpenAI said all three new AI models can “independently accomplish tasks on behalf of users.” The GPT‑ 4.1, GPT‑ 4.1 mini, and GPT‑ 4.1 nano will be available to developers via the Application Programming Interface (API).

Also read: UIDAI unveils new Aadhaar app for instant face-based ID verification - All details

Now, with a more stable model launching, OpenAI will slowly remove the GPT-4.5 preview mode by July 14. With new powerful AI models paving their way, OpenAI is all set to compete with other tech giants such as Google and Anthropic by competing with the latest AI models. Furthermore, the company is also working to introduce the anticipated GPT-5 model which could officially debut this year.