Enterprise AI adoption is moving in a clear direction: Anthropic has become the leading choice for large language model usage among businesses, according to a report from Menlo Ventures released on Thursday. The change puts Anthropic ahead of OpenAI in enterprise usage, reversing the picture from just two years ago.
The shift matters because enterprise usage is a different test from consumer popularity. Businesses tend to evaluate models through daily workflows, developer adoption, coding performance, and whether a model fits the practical demands of teams using AI at work.
Anthropic has moved ahead in enterprise AI usage
Anthropic now holds 32% of the enterprise large language model market share by usage, according to the Menlo Ventures report. OpenAI is second with 25%.
That is a major reversal from the earlier enterprise market. In 2023, OpenAI held 50% of enterprise market share by usage, while Anthropic had 12%. Since then, the report says OpenAI’s enterprise share has declined sharply, while Anthropic’s has steadily risen.
Google has also seen enterprise usage for its models increase over the last few years. The source does not give Google’s specific share, but it places Google within a broader trend: enterprise AI usage is no longer centered around one provider in the same way it was in 2023.
For companies choosing AI tools, the numbers suggest a market that is still being actively reshaped. Early leadership did not permanently settle enterprise demand. Instead, businesses appear willing to move toward models that better match the tasks they care about most.
Coding is where Anthropic’s lead is widest
The strongest signal in the Menlo Ventures findings comes from coding. Anthropic holds 42% of enterprise market share in coding, making it the largest player in that category by a wide margin.
OpenAI’s coding share is 21%, meaning enterprise usage of Anthropic’s AI models for coding is more than double OpenAI’s share. That gap helps explain why Anthropic’s broader enterprise position has strengthened: coding is one of the clearest work settings where teams can repeatedly test model usefulness.
In practical terms, coding use is not just about writing snippets. Enterprises using AI for software work are likely judging whether models can support developer workflows reliably enough to become part of daily operations. The source does not detail the specific reasons enterprises prefer one model over another, but the market share numbers show where usage has landed.
The report’s findings also line up with anecdotal chatter in the industry, which suggested that enterprise and startup developers preferred Claude over OpenAI’s ChatGPT. That does not mean every enterprise developer uses Claude, but it does show that the reported usage data reflects a pattern many in the industry had already been discussing.
Claude 3.5 Sonnet and Claude 3.7 Sonnet shaped the momentum
According to the Menlo Ventures report, Anthropic’s release of Claude 3.5 Sonnet in June 2024 laid the foundation for the company’s surge in usage. The release of Claude 3.7 Sonnet in February 2025 then accelerated that momentum.
Those releases are important because they connect the market share shift to specific product moments named in the source. Anthropic’s rise was not presented as a sudden unexplained jump. The report ties it to the performance and adoption cycle around Claude 3.5 Sonnet and Claude 3.7 Sonnet.
For enterprise buyers, model selection is often shaped by visible improvements that arrive through new releases. When a model becomes more useful for core work, teams can begin to shift usage. If that usage keeps expanding across organizations, market share can change quickly.
OpenAI remains highly significant, but the enterprise numbers show a different balance than the consumer conversation might suggest. OpenAI reported last week that its users send more than 2.5 billion prompts to ChatGPT a day, giving it a strong consumer foothold. Enterprise usage, however, is where Anthropic is now ahead in the Menlo Ventures report.
Closed models remain the enterprise preference
The report also found that enterprises prefer closed models, the kind used by Anthropic and OpenAI. More than half of enterprises replied that they do not use open source models at all.
Open source usage has also fallen in daily enterprise workloads. Only 13% of enterprise daily workloads use open source models as of mid-year 2025, down from 19% at the beginning of the year.
Meta still maintains dominance in the open source market. Even so, the enterprise workload numbers in the source point toward a narrower role for open source models in day-to-day business use than at the beginning of the year.
The finding is notable because open source models are often discussed as an important alternative to closed systems. In the enterprise data cited here, however, closed models are where most usage is concentrated. That gives companies like Anthropic and OpenAI a strong position in the business market, even as their relative shares shift.
What the enterprise AI shift means
The Menlo Ventures report presents a market where enterprise AI leadership is fluid. OpenAI had 50% of enterprise usage share in 2023, but Anthropic now leads with 32%. In coding, Anthropic’s 42% share gives it an even clearer advantage.
The broader implication is that enterprise AI adoption is being shaped by use cases, not just brand recognition. Consumer usage and business usage can point in different directions. OpenAI has a strong consumer footprint through ChatGPT, while Anthropic has become the top enterprise choice by usage in this report.
Several facts stand out from the source:
- Anthropic holds 32% of enterprise large language model market share by usage.
- OpenAI holds 25% of enterprise market share by usage.
- Anthropic holds 42% of enterprise market share in coding.
- OpenAI holds 21% of enterprise market share in coding.
- Only 13% of enterprise daily workloads use open source models as of mid-year 2025.
For the AI market, the message is straightforward. Enterprise users are making different choices than they did in 2023, and Anthropic has become the main beneficiary of that change.