Try the Model

Use the widget below to experiment with GPT-4.1. You can detect COCO classes such as people, vehicles, animals, household items.

Overview

GPT-4.1 is a new, multimodal model developed by OpenAI. GPT-4.1 models have a context window of 1 million tokens, making the models ideal for tasks that require long context. The models achieve state-of-the-art performance on a range of benchmarks, including SWE-bench Verified, Video-MME, and Scale’s MultiChallenge benchmark.

GPT-4.1 comes in three sizes:

GPT-4.1
GPT-4.1 mini
GPT-4.1 nano

The smaller the model, the lower the latency you can expect from the model.

GPT-4.1 is only available in the OpenAI API. According to the model launch post, the model “will only be available via the API”. The reasoning cited is:

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.

You can also experiment with the model in the ChatGPT Playground, ideal for exploring the model capabilities in a visual interface.

GPT-4.1 License

GPT-4.1

is licensed under a

license.

Performance

Deploy a GPT-4.1 API

You can use Roboflow Inference to deploy a

GPT-4.1

API on your hardware. You can deploy the model on CPU (i.e. Raspberry Pi, AI PCs) and GPU devices (i.e. NVIDIA Jetson, NVIDIA T4).

Below are instructions on how to deploy your own model API.

Label Data Automatically with GPT-4.1

You can automatically label a dataset using GPT-4.1 with help from Autodistill, an open source package for training computer vision models. You can label a folder of images automatically with only a few lines of code. Below, see our tutorials that demonstrate how to use GPT-4.1 to train a computer vision model.

No items found.