In a major advancement for enterprise AI accessibility, Amazon Web Services (AWS) has announced the integration of OpenAI’s open-weight models—gpt-oss-120b and gpt-oss-20b—into Amazon Bedrock and Amazon SageMaker AI. This marks the first time OpenAI’s models are directly accessible through AWS’s AI services, allowing thousands of organizations to seamlessly build and scale generative AI applications with advanced reasoning, agentic capabilities, and coding intelligence.
The gpt-oss-120b model, in particular, stands out for delivering up to 3x better price-performance than other large models like Gemini and DeepSeek, making it an attractive choice for high-scale enterprise use.
“Open weight models are an important area of innovation… and the addition of OpenAI as our newest open weight model provider marks a transformative shift in access to advanced technology,” said Atul Deo, Director of Product at AWS.
Echoing this, Dmitry Pimenov from OpenAI emphasized the customer impact: “Together with AWS, we’re providing powerful, flexible tools that make it easier than ever for customers to build, innovate, and scale.”
These models come with robust built-in safety controls, full compatibility with Amazon Bedrock AgentCore, and a 128K context window, enabling customers to work with extended documents, research papers, and long transcripts. They are ideal for a wide range of high-value use cases—from scientific modeling and complex reasoning to enterprise-grade agentic workflows.
Customers can also enhance model usage by integrating features such as Guardrails for safety, Custom Model Import, and Knowledge Bases—enabling highly tailored and secure AI experiences.
Already in use by global enterprises like Pfizer, GoDaddy, Siemens, and Lonely Planet, Amazon Bedrock continues to evolve as a trusted foundation for deploying responsible AI at scale—now with more flexibility and power through OpenAI’s latest contributions.
 has announced the integration of OpenAI’s open-weight mod...){kind=link}