Top News

Technology Desk: After a long wait and teaser, OpenAI has launched its new AI model GPT-6 Astra. The company claims that Astra is its most powerful AI model till date. According to OpenAI, this model can be a big step towards AGI i.e. Artificial General Intelligence in the world of AI.

According to benchmarks shared by OpenAI, GPT-6 Astra has outperformed other advanced AI models in the market in many cases. The company claims that it is even ahead of Anthropic's Claude Fable 5.1.

Sam Altman announces the launch

OpenAI CEO Sam Altman posted on social media platform X and informed about the launch of the new model. He said that GPT-6 Astra is the world's best model in many areas including computer use, professional work, science, coding and cyber security. According to Altman, this model can help in creating new types of entrepreneurship, scientific discoveries and new things.

Is Astra the beginning of AGI?

OpenAI President Greg Brockman made a big claim about GPT-6 Astra in a media briefing. He said that it would not be wrong to think that the world has now reached the era of AGI.

Brockman believes that in the years to come, people may look back on Astra's launch as a turning point in the history of AI. According to him, in the future when people will look back and ask when AGI actually came, it is possible that they will consider this era and this model as its beginning.

What is AGI?

AGI i.e. Artificial General Intelligence is called such AI technology, which is capable of understanding and doing different types of tasks like humans. Existing AI models are generally good at certain tasks, while AGI aims to perform a variety of tasks with human-level understanding.

However, there is no single universally accepted definition of AGI and there is ongoing debate among experts as to whether an AI model should be considered AGI or not.

Claims better performance in coding too

OpenAI has described GPT-6 Astra as its best model yet for software engineering. According to the company, Astra performed better in tasks like finding bugs, completing terminal- tasks and answering questions to large codebases.

According to OpenAI, Astra outperformed its older GPT-5.6 Sol and Anthropic's Fable models in these benchmarks. The company says that the use of Astra will not be limited to just chat or general questions and answers. It can also be used in areas like working with computers, professional tasks, research, coding and cyber security.