AI research lab Anthropic’s AI models are now the top choice among enterprises, surpassing OpenAI. according toMenlo VenturesAnthropic now holds 32% of the enterprise large-scale language model market by usage, according to a report released Thursday. OpenAI is in second place with a 25% share.
This figure represents a significant reversal from a few years ago. The report shows that OpenAI’s market share among enterprises has declined sharply since 2023, while Anthropic’s market share has steadily increased during the same period. Two years ago, OpenAI held 50% of the enterprise market, while Anthropic had only 12%.
Google has also seen an increase in enterprise usage of its model over the past few years.
In the programming world, Anthropic has an even bigger market share, with a commanding 42% of the enterprise market share. In the coding space, enterprise-level usage of Anthropic AI models is more than double that of OpenAI, which holds 21% of the overall market share.
Anthropic released the Claude 3.5 Sonnet model in June 2024, setting the stage for a surge in usage from the company, the report said. The release of Claude 3.7 Sonnet in February 2025 has accelerated this momentum.
Menlo Ventures' findings are consistent with industry anecdotes that enterprises and startup developers prefer Claude over OpenAI's ChatGPT. Meanwhile, OpenAI is also firmly entrenched in the consumer space. The company reported last week that its users send more than 2.5 billion tips to ChatGPT every day.
The Menlo Ventures report found that companies prefer closed models, which are also used by Anthropic and OpenAI. More than half of businesses say they don't use an open source model at all. As of mid-2025, only 13% of daily enterprise workloads use an open source model, down from 19% at the beginning of the year. Meta still dominates the open source market.