A new Anthropic AI model could arrive sooner than expected as the company faces growing competitive pressure from OpenAI’s GPT-6 Astra.
According to a September 19 report from Reuters, Anthropic is considering rolling out another AI model as it weighs how to respond to Astra’s growing momentum with businesses and developers.
No new Anthropic model has been officially announced.
There is no confirmed model name.
There is no confirmed release date.
But the discussions are notable for another reason.
Only days earlier, Anthropic CEO Dario Amodei called on the AI industry to deliberately slow the pace at which frontier models become more capable.
Now competitive pressure may be pushing Anthropic in the opposite direction.
Anthropic Is Reportedly Weighing an Earlier Model Release
Reuters reported that Anthropic is considering releasing a new model partly in response to the momentum OpenAI has gained since introducing GPT-6 Astra.
The report cited three people familiar with the matter.
Anthropic declined to comment.
That means this should not be treated as a confirmed product launch.
The company could release the model, change its timing or decide not to launch it yet.
One source told Reuters that Anthropic is evaluating the safety of its next model as part of the decision.
That detail is important because Anthropic has spent years positioning itself as one of the most safety-focused companies in frontier AI.
The company has built much of its identity around responsible scaling, model evaluations and safeguards designed to become stronger as AI capabilities increase.
Just this month, Anthropic has continued expanding that work.
The company recently published new measurements for tracking the pace of AI development and announced a partnership with Accenture involving independent evaluation and red-teaming of frontier models.
At the same time, Anthropic is competing in one of the fastest-moving technology markets in the world.
And OpenAI just changed the competitive landscape.
GPT-6 Astra Is Putting New Pressure on Anthropic
OpenAI released GPT-6 Astra on September 3.
OpenAI describes Astra as its most capable broadly available model yet, with improvements in computer use, browsing, software engineering, cybersecurity, science and professional work.
The company is particularly emphasizing Astra’s ability to perform complicated work across multiple steps rather than simply answer individual prompts.
That matters in the enterprise market.
Businesses increasingly want AI systems that can do more than generate text.
They want models that can research information, work with software, analyze data, write code, use business applications and complete workflows with less supervision.
Those are also areas where Anthropic has been aggressively expanding Claude.
We recently covered Anthropic’s own business push in Claude for Small Business Gets a Major Upgrade: 43 Workflows and 27 New Integrations.
The competition is therefore no longer just about which chatbot gives the better answer.
OpenAI and Anthropic increasingly want their AI systems embedded inside the work businesses perform every day.
Astra Is Already Showing Signs of Enterprise Momentum
The early numbers help explain why Anthropic is reportedly paying attention.
Reuters reported that GPT-6 Astra represented about 13% of enterprise AI spending tracked by Ramp, compared with roughly 8% for Anthropic’s Claude Fable.
That does not mean Astra has 13% of the entire enterprise AI market.
It reflects spending observed through Ramp’s data and should be interpreted within that dataset.
Still, it is an unusually fast showing for a model released only this month.
Another signal comes from OpenRouter.
OpenRouter routes developer requests across models from different AI companies, making its usage patterns one useful indicator of what developers are actually choosing.
According to Reuters, users on OpenRouter spent more on OpenAI models than Anthropic models during the previous week.
It was reportedly the first time OpenAI had led Anthropic on that measure in more than two and a half years.
For a market that can change quickly with every major model release, that shift matters.
Anthropic Still Has a Major Enterprise Position
None of this means Anthropic has suddenly lost the enterprise AI race.
Far from it.
Anthropic remains deeply established with businesses.
Reuters reported that Anthropic’s annualized revenue run rate had surpassed $65 billion by the end of July, compared with more than $40 billion for OpenAI.
An annualized run rate is not the same as annual profit or even completed annual revenue. It generally projects the current pace of revenue over a full year.
But the figures illustrate the scale Anthropic has reached.
Separate Ramp data also continues to show Anthropic as one of the most widely adopted foundational AI providers among businesses in its dataset.
Enterprise customers can also be difficult to displace once they have built a model into internal workflows, applications and company processes.
Changing AI providers can require new integrations, testing, security reviews, employee training and procurement approvals.
That gives established providers some protection from competitors.
But it does not eliminate the pressure to keep improving.
A powerful new model can change buyer expectations almost overnight.
And that creates an uncomfortable situation for Anthropic.
