OpenAI has publicly released its flagship artificial intelligence technology, GPT-5.6 Sol, its most powerful model yet, alongside a new tool to assist with everyday office work.
The new release, on Thursday, escalates the battle for A.I. dominance with the rival start-up Anthropic as the technology grows increasingly sophisticated.
Along with its new model, OpenAI unveiled a new tool, ChatGPT Work, which is powered by Sol. Much like Anthropic’s Claude Cowork, ChatGPT Work is an “A.I. agent” that can operate software apps and websites on behalf of users, including spreadsheets, online calendars and email services.
When running standard benchmark tests, OpenAI’s new flagship model roughly matches the performance of Anthropic’s leading model, Fable 5, according to Vals AI, a company that tracks the performance of the latest A.I. technologies.
“In real-world situations involving financial and legal tasks, it is state of the art,” said Rayan Krishnan, the chief executive of Vals AI.
The release of the more powerful model followed U.S. government efforts to restrict both OpenAI’s and Anthropic’s new A.I. models over cybersecurity concerns. While President Trump had long taken a hands-off approach in regulating the A.I. industry, he signed an executive order last month that asked tech companies to voluntarily give the government oversight of new A.I. models before releasing them to the public.
Last week, OpenAI said it would share its new technology with only a small group of companies approved by the Trump administration. But after lengthy discussions, the administration allowed both OpenAI and Anthropic to release their new technologies publicly.
Although the administration initially treated the technologies as threats to national security, many independent security experts said the dangers were overstated.
OpenAI is taking a more open approach to cybersecurity than Anthropic, which has tightly restricted the use of Fable to prevent hackers from using its technology. OpenAI’s new model includes fewer guardrails, so that individuals and businesses can use the technology to defend themselves.
Anthropic released a new A.I. system called Claude Mythos in April, but said it couldn’t share the technology with the public. Some executives and government officials worried that the technology would usher in a new era of cybersecurity threats.
It released Fable 5, a more limited version of Mythos, last month. Days later, the Trump administration ordered Anthropic to suspend access for all foreign nationals, citing national security concerns. To comply with the order, Anthropic took the technology offline entirely.
The company has added guardrails to Fable designed to block responses related to cybersecurity, biology and other vulnerable areas. When Fable blocks responses, it sends them to a less powerful technology, Claude Opus 4.8.
Technologies like Fable and Sol are particularly good at identifying vulnerabilities in the software that underpins the internet. This means that malicious hackers can use the technology to attack computer networks. But it also means that businesses and governments can use it to defend against these attacks.
Fable’s guardrails may prevent hackers from using the technology for offensive purposes. But the same guardrails may also bar businesses and governments from using the system for defense.
Like Anthropic’s Fable, OpenAI’s GPT-5.6 Sol is expensive compared with less powerful models. Fable is roughly twice as expensive as Anthropic’s next most powerful model, and OpenAI’s model is more expensive than Fable in some cases, Mr. Krishnan said.
On Thursday, Meta also released a new A.I. technology. It was not as powerful as GPT-5.6 Sol or Fable, and is also considerably less expensive.
(The New York Times has sued OpenAI and Microsoft, claiming copyright infringement of news content related to A.I. systems. The two companies have denied those claims.)
