Logo La Derechadiario US
ES
Logo X
OpenAI logo on the left and a person holding a mobile phone with an internet connection error message on the right
OpenAI introduces two offline AI models for PC and mobile | Camara La Derecha Diario
ARGENTINA

OpenAI launches two free ChatGPT models that work without the internet

OpenAI introduced two open-source AI models that work without internet access and already outperform the successful DeepSeek R1

by

Editorial Team

OpenAI introduced two open-source artificial intelligence models that work without needing to be connected to the internet. They're available for installation on computers, laptops, or cell phones.

This is a direct response to pressure from DeepSeek and its successful R1 model. With this move, OpenAI is once again competing strongly in the open-source field.

Person using ChatGPT on a mobile phone while a screen with the same application is in the background
OpenAI introduced two open-source artificial intelligence models | La Derecha Diario

Offline AI: Which models did OpenAI launch?

The models are called gpt-oss-120b and gpt-oss-20b. The first has greater capacity and is designed for PCs with NVIDIA GPU, while the second is lighter and even works on laptops with 16 GB of RAM.

Both can be installed locally and don't require an internet connection. This is an important step toward more private, offline AI.

Where can they be downloaded?

  • From Hugging Face (for Ollama, updated interface).
  • From NVIDIA's website (with ready-to-use web version).
Minimalist chat window with a drawing of a llama in the center and a field to send messages at the bottom
Hugging Face (for Ollama, updated interface) | La Derecha Diario

How do they perform compared to other models?

They outperform the DeepSeek R1 model. In Codeforces benchmarks, they reach scores of 2622 and 2516 points. Although they don't match models like o3 or o4-mini, they offer excellent performance in open-source AI.

In the Last Exam of Humanity test, their scores were 19% and 17.3%. They also stand out in tasks such as tool invocation, code execution, and web search.

gpt-oss performance
How do they perform compared to other models? | La Derecha Diario

What about hallucinations?

The models have a high level of hallucinations: they generate up to three times more incorrect answers than the o1 model. OpenAI hasn't identified the cause yet.

It's always recommended to verify sources if they're used in sensitive or analytical tasks.

A strategic move against the competition

With these launches, OpenAI aims to compete with the progress of open models created in China. Additionally, it breaks with its historic closed model, although it keeps training data secret.

Screenshot of a tweet mentioning that gpt-oss is a state-of-the-art reasoning model with open weights, performance comparable to o4-mini, and can run locally, considering it the best and most usable open model in the world.
OpenAI seeks to compete with the advancement of open models developed in China | La Derecha Diario

Both models are licensed under Apache 2.0, which allows unrestricted use for research, customization, or commercial development.

➡️ Argentina

More posts: