Amazon Web Services (AWS) has made a big announcement. Now OpenAI’s open weight AI models are available directly on the AWS platform. For the first time, developers, companies and general users who use AWS can now directly use OpenAI’s powerful models. The same OpenAI that created ChatGPT.
The names of these models are gpt-oss-120b and gpt-oss-20b. These models are designed for complex tasks such as answering questions, solving mathematical problems, writing code and analyzing scientific data. The special thing is that these models are open weight, that is, they can be easily used and can also be changed as per the need. Usually other AI systems are not so open.
By bringing OpenAI’s models to AWS, Amazon has made advanced AI tools accessible to more and more people. Now no one will need a supercomputer to use them. These models can be easily run on AWS, which is the world’s most used cloud platform.
These models can be used in two main services of AWS:
- Amazon Bedrock: This is a service where you can build AI applications without worrying about servers. The large OpenAI model in it, gpt-oss-120b, is very efficient. It performs better than Google (Gemini), DeepSeek and OpenAI’s older models and works at a lower cost.
- Amazon SageMaker: This is for those who want more control. In this, you can train, test, tune and deploy the model yourself as you want.
Now the question is who can use it? The answer is – anyone. From startups to big companies. Companies like Pfizer, Lonely Planet, Siemens, and even the PGA TOUR are already using Amazon Bedrock to build smart apps and improve their services. Now that OpenAI models have been added to the mix, these companies will be able to make even more progress.
These models are especially useful for AI agents. You can think of them as smart digital assistants that can read documents, understand tasks, and act on complex instructions. They are great for customer service, research, coding, and understanding legal or technical documents.
One of the features of these models is that they can read very long inputs. Their input window is up to 128K tokens — this means they can understand long content like entire reports, chat records, or research papers without losing context.
Talking about security, OpenAI has trained these models in such a way that they do not produce harmful or incorrect content. AWS also offers tools like Guardrails that filter out false positives, making it safer for businesses.
In essence, this means that more and more people can now use the world’s most advanced AI tools — without having to be an AI expert. OpenAI’s models are now available on AWS alongside other major AI providers like DeepSeek, Meta, and Mistral AI. This gives AWS users more options than ever before.
Today, there are more than 100 AI models available on Amazon Bedrock. This means that AWS is now a major hub for anyone who wants to use AI technology at scale. This move could accelerate innovation across different industries and make it easier than ever to bring AI into everyday tools and apps.
