Anthropic Gains AI Market Share as Enterprise Demand Accelerates, Menlo’s Matt Murphy Says

Anthropic continues to gain ground in the artificial intelligence market as businesses spend more on advanced models, according to Matt Murphy, a partner at Menlo Ventures, an early investor in the company.

Speaking on CNBC, Murphy said Anthropic’s third-quarter revenue rose approximately 45% from the previous quarter to $16.8 billion. The growth comes as the company considers a potential initial public offering as early as November, according to the report.

Murphy challenged concerns that AI demand is weakening, citing Menlo’s enterprise research as evidence that spending continues to accelerate. The firm’s tracking indicates the overall enterprise generative AI market is growing roughly fivefold this year, compared with 3.2 times in 2025, he said.

His assessment also challenges the view that rising usage of open-weight models is necessarily eroding the business of proprietary developers such as Anthropic and OpenAI. Open-weight models make their underlying parameters available, giving businesses more flexibility to customize and deploy them.

Murphy addressed reports that open-weight models had increased their share of token volume from roughly 20% to 80% over a 12-week period. Tokens measure the units of information processed by AI systems, but their volume does not directly translate into revenue or customer spending.

Measured by revenue, Murphy said, demand remains strong for the most capable proprietary models. He cited OpenRouter, a marketplace that provides access to multiple AI models, where he said Anthropic captured more than 70% of both token volume and revenue following a Claude release. Those figures describe activity on that marketplace rather than the broader AI industry.

Businesses are testing open-weight alternatives to control costs, Murphy said, but he expects most companies to limit those models to approximately 20% to 30% of their token usage. His argument is that more demanding applications still require the capabilities of leading proprietary systems.

Menlo’s investment in Anthropic gives the firm a financial interest in the company’s success. Even so, Murphy’s comments highlight a central question for AI investors: whether lower-cost models will capture spending alongside usage, or whether customers will continue paying a premium for stronger performance.

Murphy also said he expects both Anthropic and OpenAI eventually to enter the public markets. He described their growth as exceptional, drawing on his more than 25 years in technology investing, and argued that Anthropic has built a stronger enterprise business.

The timing of any offering remains uncertain. For investors, the key measures will include revenue growth, the durability of enterprise demand and how effectively each company converts model usage into paying customers.