OpenAI has introduced two new open-weight language models designed for advanced reasoning that are small and efficient enough to run directly on personal laptops. These models, gpt-oss-120b and gpt-oss-20b, match the performance of OpenAI’s smaller proprietary reasoning tools and offer developers the freedom to operate them on their infrastructure.
Open-weight models differ from open-source models in a key way: while they share the trained weights of the model for fine-tuning and analysis, they don’t include full access to training data or source code. OpenAI co-founder Greg Brockman emphasized the practical value of this approach, saying, “One of the things that is unique about open models is that people can run them locally. People can run them behind their firewall, on their infrastructure.”
How Are These Models Being Used and Distributed?
Amazon Web Services (AWS) is now distributing OpenAI’s models via its Bedrock generative AI marketplace, marking the first time OpenAI has been featured there. “OpenAI has been developing great models, and we believe that these models are going to be great open-source options, or open-weight model options for customers,” said Atul Deo, director of product at AWS Bedrock.
This release signals a competitive shift in the AI ecosystem. Meta’s Llama models once dominated the open-weight category, but recent breakthroughs from China’s DeepSeek challenged that position. Now, OpenAI re-enters the open-model race for the first time since releasing GPT-2 in 2019.
What Makes These New Models Stand Out?
The gpt-oss-120b model can run on a single GPU, while the smaller gpt-oss-20b is optimized for consumer-level machines. Both models perform particularly well in coding, math competitions, and health-related queries. They were trained on a text-only dataset that included general knowledge with a focus on math, science, and coding.
Despite their strong capabilities, OpenAI did not provide direct performance comparisons with rival models like DeepSeek-R1. However, they claim parity with their in-house models o3-mini and o4-mini.
OpenAI’s move comes amid a broader industry push for more flexible, transparent AI tools. The company, backed by Microsoft and currently valued at $300 billion, is reportedly seeking an additional $40 billion in funding, led by Softbank.
By Risper Akinyi