In the market for large language model (LLM)-based AI, Anthropic has been called part of a Big 3 with OpenAI and Google, but it has often carried a strong No. 3 image. That is changing. Anthropic has sharply raised its standing in the tech world since the start of the year. A flood of AI industry headlines has featured Anthropic in recent days. Many have carried the view that Anthropic has overtaken OpenAI and Google.

The shift in mood accelerated after Anthropic unveiled Claude Co-Work, an AI tool, in January. Claude Co-Work is an AI tool that lets non-developers build work tools, and it is still in preview. Even so, a narrative that it could threaten the existing software market has helped put Anthropic in the ranks of AI agenda setters beyond OpenAI and Google. After Claude Co-Work was unveiled, shares of major software companies fell sharply, and at least in stock markets the threat from Anthropic became reality. That has driven a surge in articles about Anthropic's impact across the broader economy, beyond AI and tech. In the process, the phrase “death of software” became widely circulated.

Anthropic's latest model, Claude Opus 4.6, unveiled on Feb. 5 local time, also helped spread the “death of software” narrative further. After Claude Opus 4.6, shares of software companies such as Salesforce and Intuit fell again.

According to the company, Claude Opus 4.6 focuses on new methods of model training and interaction with applications and databases. Claude Opus 4.6 analyses and synthesises data, runs a coding assistant team, and also provides functions similar to product management.

Dean Ball (딘 볼), a senior fellow at the Foundation for American Innovation who publishes an artificial intelligence newsletter, told the Wall Street Journal, “The popular spread of word-of-mouth about Anthropic models is the most important event in AI since the launch of ChatGPT,” adding, “It’s very interesting.”

Anthropic, now in its fifth year, has emphasised an LLM strategy targeting the enterprise market (B2B) and developers. Compared with OpenAI and Google, which also cover individual users, that is why it drew relatively less attention. But the story changes if the focus is on substance rather than awareness. A strategy of selection and focus led to Anthropic securing a clear advantage in the enterprise and developer markets.

The numbers speak. A recent Financial Times report said Anthropic's annualised revenue rose from $1 billion in early last year to more than $9 billion by the end of last year. Anthropic told investors that annualised revenue would exceed $30 billion this year and continue to rise sharply after that. The FT, citing 12 investors, reported that enterprise market share, a product strategy with a clear message and stable leadership are making Anthropic increasingly seen as a safer long-term bet than OpenAI.

According to Silicon Valley venture investment firm Menlo Ventures, in the first half of 2025 Anthropic overtook OpenAI to rank first in enterprise LLM API usage. Anthropic recorded 32 percent of the overall market, ahead of OpenAI at 25 percent. Google followed with 20 percent and Meta with 9 percent. As recently as the end of 2023, a year and a half earlier, OpenAI accounted for half of the enterprise LLM API market, but it failed to maintain its momentum.

Data from cost management startup Ramp, cited by the WSJ, show that as of January Anthropic led in share of API spending at 80 percent. API costs are incurred when users access models through external services.

Anthropic's presence stands out in the AI coding tools market. Anthropic's Claude Code entered a $1 billion annual revenue track about a year after its launch. According to one partner at Menlo Ventures, which first invested in Anthropic in 2023, Anthropic initially developed Claude Code for internal use but moved aggressively to productise it because it was so good.

Claude Code also served as a spearhead for Anthropic in gaining an advantage in the enterprise market.

Software engineering was a key factor in attracting corporate customers, and Anthropic executives expected this to ultimately serve as a link to expanding into other areas. That was because a model proficient in coding can perform almost any task on a computer.

According to the WSJ, David Hsu (데이비드 슈), CEO of Retool, a startup that builds and manages enterprise AI tools, said, “Anthropic has been targeting the enterprise market, and it realised that in the end the key is coding.”

Anthropic is also close to raising $35 billion in funding at a valuation of $350 billion, and it is also aiming for an initial public offering (IPO) this year, becoming the first LLM development startup to do so. It is likely to raise investment at the level it is considering, and the mood around an IPO could change depending on how much it maintains its current momentum.

With Google and OpenAI also accelerating their push into coding AI and the enterprise market, if Anthropic maintains or strengthens its edge, Anthropic's box-office power in the AI arena is likely to grow further. This year looks set to be a very important one for Anthropic.