"Anthropic Weighs Early Release of New Model Amid Astra's Rapid Rise"

As OpenAI's recently launched GPT-6 Astra makes rapid strides in the enterprise market, Anthropic is considering accelerating the release of a new artificial intelligence (AI) model ahead of its initial public offering (IPO), Reuters reported, citing three sources.

The plan to release a new model before an IPO comes after AI company leaders, including Anthropic Chief Executive Officer (CEO) Dario Amodei, recently emphasized AI safety and urged a slower pace.

The rollout of this new model is expected to be a test of whether Anthropic can maintain its safety-first identity, which has set it apart from competitors in its rivalry with OpenAI.

OpenAI launched GPT-6 Astra earlier this month, stating that its performance has improved in computer use, software engineering, cybersecurity, and specialized tasks.

The model has drawn a strong response from enterprise users and developers.

According to corporate spend management platform Ramp, Astra's share of enterprise AI spending reached about 13%, surpassing Anthropic's Claude Feble at around 8%.

On OpenRouter, which connects developer traffic to AI models, spending on OpenAI models surpassed Anthropic last week for the first time in two and a half years.

Investors reviewing potential investments in the IPO are closely watching whether Anthropic, long considered a front-runner in the enterprise AI market, could lose market share to OpenAI.

Investors expect the leading position to shift repeatedly as Anthropic, OpenAI, Google, and other major AI developers roll out new generations of models.

Any advantage could be short-lived.

Sources said discussions are also underway to balance investment for new model releases with efforts to strengthen corporate profitability, as rising interest rates cause investors to focus on the timing of returns.

Investors noted that the rise of open-source and open-weight models could pose a greater challenge to major AI companies such as Anthropic.

Such models could lower token costs and enable companies to build more of their own AI infrastructure.

Previously, The Wall Street Journal (WSJ) reported that Anthropic is pushing for an IPO in November, later than initially expected by the market.

OpenAI reaffirmed on the 19th that it will not go public within this year.

CEO Altman stated that given concerns over AI safety, now is not the right time to go public.
※ Please note: This article was translated by AI and may contain errors.